SPE 70067

A Strategy for Attacking Excess Water Production

R. S. Seright, SPE, New Mexico Petroleum Recovery Research Center
R. H. Lane, SPE, Northstar Technologies International
R. D. Sydansk, SPE, Sydansk Consulting Services

Copyright 2001, Society of Petroleum Engineers Inc.

to be 5-10 billion dollars in the US and around 40 billion
This paper was prepared for presentation at the SPE Permian Basin Oil and Gas Recovery dollars worldwide.2
Conference held in Midland, Texas, 15–16 May 2001.
Many different causes of excess water production exist
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
(Table 1). Each of these problems requires a different
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to approach to find the optimum solution. Therefore, to achieve a
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at high success rate when treating water production problems,
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
the nature of the problem must first be correctly identified.3
for commercial purposes without the written consent of the Society of Petroleum Engineers is Many different materials and methods can be used to attack
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 excess water production problems. Generally, these methods
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.
can be categorized as chemical or mechanical (see Table 2).
Each of these methods may work very well for certain types of
problems but are usually ineffective for other types of
Abstract problems. Again, for effective treatment, the nature of the
This paper describes a straightforward strategy for diagnosing problem must first be correctly identified.
and solving excess water production problems. The strategy Four problem categories are listed in Table 1 in the general
advocates that the easiest problems should be attacked first order of increasing treatment difficulty. Within each category,
and diagnosis of water production problems should begin with the order of listing is only roughly related to the degree of
information already at hand. A listing of water production treatment difficulty. Category A, "Conventional" Treatments
problems is provided, along with a ranking of their relative Normally Are an Effective Choice, includes the application of
ease of solution. water shutoff techniques that are generally well established,
Conventional methods (e.g., cement, mechanical devices) utilize materials with high mechanical strength, and function
normally should be applied first to treat the easiest problems— in or very near the wellbore. Examples include Portland
i.e., casing leaks and flow behind pipe where cement can be cement, mechanical tubing patches, bridge plugs, straddle
placed effectively and for unfractured wells where packers, and wellbore sand plugs.
impermeable barriers separate water and hydrocarbon zones. A few comments may be helpful to clarify some of the
Gelant treatments normally are the best option for casing leaks listings in Table 1. First, the difference between Problems 1
and flow behind pipe with flow restrictions that prevent and 4 is simply a matter of aperture size of the casing leak and
effective cement placement. Both gelants and preformed gels size of the flow channel behind the casing leak. Problem 1,
have been successfully applied to treat hydraulic or natural involving casing leaks without flow restrictions, is where the
fractures that connect to an aquifer. Treatments with leak is occurring through a large aperture breach in the piping
preformed gels normally are the best option for faults or (greater than roughly 1/8 in.) and a large flow conduit (greater
fractures crossing a deviated or horizontal well, for a single than roughly 1/16 in.) behind the leak. The use of Portland
fracture causing channeling between wells, or for a natural cement is favored for treating Problem 1. Problem 4, involving
fracture system that allows channeling between wells. Gel casing leaks with flow restrictions, is where the leak is
treatments should not be used to treat the most difficult occurring through a small aperture breach (e.g., “pinhole” and
problems—i.e., three-dimensional coning, cusping, or tread leaks) in the piping (less than roughly 1/8 in.) and a
channeling through strata with crossflow. small flow conduit (less than roughly 1/16 in.) behind the leak.
The use of gel is favored to successfully treat Problem 4. In
Introduction this paper, the gels under discussion may include those formed
On average in the United States, more than seven barrels of from (1) chemically crosslinking water-soluble organic
water are produced for each barrel of oil.1 Worldwide, an polymers, (2) water-based organic monomers, or (3) silicates.
average of three barrels of water are produced for each barrel
of oil.2 The annual cost of disposing of this water is estimated
2 R.S. SERIGHT, R.H. LANE, AND R.D. SYDANSK SPE 70067

The difference between Problems 2 and 5 is again simply a should be attacked first and (2) diagnosis of water production
matter of aperture size of the flow channel behind the pipe. problems should begin with information already at hand. To
Problem 2, involving flow behind pipe without flow implement this strategy, a prioritization of water production
restrictions, is where the fluid flow is occurring through a problems is needed. Based on extensive reservoir and
large aperture flow conduit behind the pipe (greater than completion engineering studies and analyses of many field
roughly 1/16 in.). The use of Portland cement is favored to applications, the various types of water problems were
treat Problem 2. This problem is often manifested by a total prioritized and categorized from least to most difficult. This
lack of primary cement behind the casing. Problem 5, prioritization is listed in Table 1. The first three listings are the
involving flow behind pipe with flow restrictions, is where the easiest problems (Category A, Problems 1-3), and their
flow behind pipe is occurring through a small aperture flow successful treatment has generally been regarded as relatively
conduit (less than roughly 1/16 in.). The use of gel is favored straightforward. Of course, individual circumstances can be
to treat this problem. Problem 5 is often exemplified by micro- found within any of these problem types that are quite difficult
annuli flow behind the pipe. This problem often results from to treat successfully. For example, for Problem Type 3,
cement shrinkage during its curing during the well’s impermeable barriers may separate water and hydrocarbon
completion. zones. However, if many water and oil zones are intermingled
The recognition, importance, challenge, and necessity of within a short distance, it may not be practical to shut off
successfully treating Problems 2 and 5 have become much water zones without simultaneously shutting off some oil
more prominent recently with the advent of regulatory- zones. The ranking of water production problems in Table 1 is
required mechanical integrity (hydro) testing of petroleum based on conceptual considerations and issues related to the
well tubing and casing strings. ease of treating each type of problem. We realize that
Logically, identification of the excess water production operational and practical issues can make even the easiest
problem should be performed before attempting a water problems in Table 1 very difficult to solve in practice.
shutoff treatment. Unfortunately, many (perhaps most) oil and Nevertheless, the first three problem types in Table 1 are
gas producers do not properly diagnose their water production generally easier in practice to treat than the others on the list.
problems. Consequently, attempted water shutoff treatments Therefore, one should look first for these types of problems.
frequently have low success rates. Several reasons exist for the In contrast, the last three problems (Category D, Problems
inadequate diagnosis of excess water production problems. 11-13) are difficult with no easy, low-cost solution. (Gel
First, operators often do not feel that they have the time or treatments will almost never work for these problems.) The
money to perform the diagnosis, especially on marginal wells intermediate problems (Categories B and C, Problems 4-10)
with high water cuts. Second, uncertainty exists about which are caused by linear-flow features (e.g., fractures, fracture-like
diagnostic methods should be applied first. Perhaps 30 structures, narrow channels behind pipe, or vug pathways).
different diagnostic methods could be used. In the absence of a Certainly, much work remains to optimize the treatment of
cost-effective methodology for diagnosing water production these problem types. However, substantial theoretical,
problems, many operators opt to perform no diagnosis. Third, laboratory, and field progress has been made in recent years
many engineers incorrectly believe that one method (e.g., toward solving these problems—especially using gels. As will
cement) will solve all water production problems or that only be discussed shortly, Problems 4-7 (Category B in Table 1)
one type of water production problem (e.g., three-dimensional normally are best solved using gelants—i.e., the fluid gel
coning) exists. Finally, some service companies incorrectly formulation before significant crosslinking occurs. Problems
encourage a belief that a “magic-bullet” method exists that 8-10 (Category C) are best solved using preformed or partially
will solve many or all types of water production problems. formed gels (i.e., crosslinking products that will not flow into
A number of excellent papers have addressed candidate or damage porous rock).
selection and various aspects of treating specific types of A key element of the proposed strategy is to look for and
excess water production problems.2-13 A common theme of solve the easiest problems in Table 1 before attempting to
many of these papers is a need for proper diagnosis of the attack the more difficult problems. In many cases, engineers
excess water production problem. However, for the reasons initially assumed that three-dimensional coning (Problem 11
mentioned above, such diagnosis is frequently not obtained. in Table 1) caused the problem, whereas a small amount of
This paper focuses on a cost-effective strategy and subsequent diagnosis and analysis revealed the true source of
methodology for diagnosing and solving excess water water production was either flow behind pipe (Problem 2) or
production problems. The objective of this paper is to provide “two-dimensional coning” through a fracture (Problem 6).
a straightforward strategy and methodology for performing This knowledge could have substantially reduced the cost of
effective problem diagnosis so the practicing engineer does solving the problem (since Problems 2 and 6 can be solved
not forego problem diagnosis and, in turn, implement with relatively low-cost methods, whereas Problem 11
ineffective water shutoff treatments. cannot). Also, by correctly identifying the problem first, the
most appropriate method can be identified and the probability
Proposed Strategy of successfully treating the problem increases significantly.
Our proposed strategy for attacking excessive water To help implement the proposed strategy, the following
production problems advocates that (1) the easiest problems questions should be addressed in the order listed:
SPE 70067 A STRATEGY FOR ATTACKING EXCESS WATER PRODUCTION 3

then the above methods comprise a list of the first diagnostic

1. Is there a problem? methods that should be considered for implementation.
2. Is the problem caused by leaks or flow behind pipe? If a problem with unrestricted casing leaks or flow behind
3. Is the problem caused by fractures or fracture-like pipe (as defined in Table 1 and the subsequent paragraphs) is
features? identified, that problem should be addressed before attempting
4. Is the matrix-flow problem compounded by crossflow? to solve additional, more difficult problems that may exist.
Some engineers disagree with this suggestion—arguing that
Is There a Problem? An important first question when they wish to apply a water shutoff method that solves multiple
attacking a water production problem is, do significant types of problems at once. While this fortuitous circumstance
volumes of mobile oil remain in the pattern or in the vicinity occasionally occurs, the optimum solution for treating the
of the well of interest? Three types of observations are different types of problems usually varies considerably. For
commonly used to make this assessment. First, a pumper may example, the optimum solution for an unrestricted flow behind
notice that certain well(s) exhibit a sudden increase in water pipe problem and that for a fracture that leads to an aquifer
cut. Second, a well or pattern of wells may be noted as may differ considerably in (1) desired properties of the
producing at significantly higher water/oil ratios (WORs) than blocking agent, (2) volume of blocking agent placed, and (3)
other similar patterns. Third, plots of fluid production versus placement method. Thus, although a chosen treatment method
time may show an abrupt increase in WOR at a certain point. may be effective in treating one of these two excessive water
Results from reservoir simulation studies constitute a fourth, production problems, the chosen treatment will most likely be
less common method sometimes used by large oil producers to ineffective in treating the other water production problem.
analyze water production problems in large reservoirs.
The oilfield operator should recognize that two distinct Is the Problem Caused by Fractures or Fracture-Like
types of water production exist. The first type, usually Features? A critical aspect of diagnosing most excess water
occurring later in the life of a waterflood, is water that is co- production problems is deciding whether fluid flow around the
produced with oil as part of the oil’s fractional flow wellbore is radial or linear. Flow behind pipe, fractures, and
characteristics in reservoir porous rock. If production of this fracture-like features are associated with linear flow, while
water is reduced, oil production will be reduced radial flow generally occurs in matrix reservoir rock when
correspondingly. The second type of water production directly these features are absent. (We recognize the special case of
competes with oil production. This water usually flows to the radial flow in fractures, e.g., for vertical fractures that cross
wellbore by a path separate from that for oil (e.g., water horizontal wells. This case will be treated separately later. In
coning or a high permeability water channel through the oil this section, our consideration of radial flow is confined to
strata). In these latter cases, reduced water production can flow in matrix, while linear flow refers to the presence of
often lead to greater pressure drawdowns and increased oil extremely permeable fracture-like features.) Simple
production rates. Obviously, the second type of water calculations using the Darcy equation reveal that the approach
production should be the target of water shutoff treatments. for solving these linear flow problems must be fundamentally
Understanding and conceptualizing the reservoir different than solving radial flow problems in matrix reservoir
“plumbing” is a key to (1) distinguishing between the above rock or sand.14 Especially for gel treatments, linear versus
two types of water production, (2) successfully diagnosing the radial flow problems differ radically in (1) gel properties
water production problem, and (3) successfully implementing desired, (2) placement procedures required, and (3) optimum
and designing water shutoff treatments. volume of the gel placed. In particular, for radial flow
problems, hydrocarbon productive zones must be protected
Is the Problem Caused by Leaks or Flow Behind Pipe? during gelant placement.14 For linear flow, an acceptable gel
Once the operator decides that the water cut is too high placement (without mechanically isolating zones) is much
considering the remaining reserves and the water is produced easier to achieve than in radial flow.
via a flow path separate from that of the oil, one should ask A number of methods are available to judge whether flow
whether the excess water production is caused by a relatively around a wellbore is linear (in fracture-like features) or radial
easy problem (as listed in Table 1)—in particular, by (in matrix rock or sand). One simple method uses the Darcy
unrestricted casing leaks or by flow behind pipe. Some of the equation for radial flow (Eq. 1).
common methods used to diagnose these problems include (1)
leak tests/casing integrity tests (e.g., hydro testing), (2) q/∆p = Σ kh / [141.2 µ ln (re /rw)] ................................... (1)
temperature surveys, (3) flow profiling tools (e.g., radiotracer
flow logs, spinner surveys, production logging tools), (4) If the actual injectivity or productivity for a well (i.e., the
cement bond logs, (5) borehole televiewers, and (6) noise logs. left side of Eq. 1, q/∆p, in BPD/psi) is five or more times
Many of these methods are used during routine surveillance greater than the injectivity or productivity calculated using the
of wells. Therefore, consistent with our proposed strategy, one Darcy equation for radial flow (i.e., the right side of Eq. 1), the
should begin the diagnostic process by examining information well probably suffers from a linear flow problem.
already at hand. If this type of information is not available,
4 R.S. SERIGHT, R.H. LANE, AND R.D. SYDANSK SPE 70067

LINEAR FLOW: q/∆p >> Σ kh / [141.2 µ ln (re /rw)] .....(2) Reportedly, these methods can estimate the fracture volume,
the fracture permeability, and, possibly under some
On the other hand, if the left side of Eq. 1 is less than or circumstances, the minimum spacing between fractures.
equal to the right side, radial flow becomes likely. Pressure interference tests can also indicate fracture
orientation. In addition to unsteady-state methods, steady-state
RADIAL FLOW: q/∆p ≤ Σ kh / [141.2 µ ln (re /rw)].........(3) productivity indexes were also suggested as a means to
estimate fracture permeability.
In the above equation, k is effective rock permeability in Interwell tracer studies provide valuable (and often
mD. If the zone contains water at residual oil saturation (Sor), k relatively inexpensive) characterizations of fractured
should take this into account. Typically the water relative reservoirs, especially for use in judging the applicability of gel
permeability at Sor is between 5% and 30% of the absolute treatments to reduce channeling.16,17 Interwell tracer data
permeability, with 10% being a good estimate if k at Sor is not provides much better resolution of reservoir heterogeneities
known. If the zone is producing only oil, k can be taken as the than pressure transient analyses.17 Tracer results can indicate
absolute permeability without incurring much error in the (1) whether fractures or fracture networks are probably present
calculation. The permeability used in Eq. 1 should be taken and if those fractures are the cause of a channeling problem,
from core analyses, log data, or pressure transient analyses. It (2) the location and direction of fracture channels, (3) the
should not be taken from production data. Net pay, h, in Eq. 1 fracture volume, (4) the fracture conductivity, and (5) the
has units of feet, while viscosity, µ, has units of cp. If the well effectiveness of a remedial treatment (e.g., a gel treatment) in
is a water injector or if the well is producing a very high water reducing channeling.18 For operators producing from mature,
cut, then the viscosity of water can be used (at the appropriate highly fractured oil reservoirs, low cost and operationally easy
temperature). If the oil cut is significant, there may be value in tracer techniques exist that can help diagnose excessive water
performing two calculations using Eq. 1, one using water production problems.
viscosity and one using oil viscosity. The natural log term in
Eq. 1 can be assumed to have a value of 6 or 7. The pressure Is the Matrix-Flow Problem Compounded by Crossflow?
drawdown or buildup (∆p, in psi) in Eq. 1 must be reasonably Once fractures and fracture-like features are eliminated as
current and applicable to the specific well of interest. It is a possibilities, the problem is deduced to be radial in nature (i.e.,
mistake to take this value from another well or to use a value radial flow exists in the matrix rock around the wellbore).
that is too old. This pressure difference indicates a great deal Next, the possibility of crossflow between reservoir strata
about the problem of the specific well and is extremely must be addressed. If fluids can crossflow between adjacent
important to measure both before and after (and even during) a water and hydrocarbon strata (and flow is radial), a gel
gel treatment. treatment should not be attempted.19 Even if gelant is only
Of course, uncertainty exists for a significant range of injected into a single zone, it will crossflow into and damage
conditions that do not satisfy either Eq. 2 or Eq. 3. Thus, the oil producing zones away from the wellbore. Thus, no
injectivity/productivity calculations will not always matter how much gelant is injected, the treatment will be
distinguish between radial and linear flow. Nevertheless, they ineffective in promoting conformance.19 In contrast, if fluids
frequently do provide a definitive indication of the flow cannot crossflow between zones and sealing Portland cement
geometry near the wellbore. Since the calculations are easily exists that prevents vertical flow immediately behind the
performed using data often at hand, they provide a low-cost casing, a gel treatment can be effective if gelant injection is
diagnostic method that should be considered when diagnosing placed only in the offending water zones.14
any excess water production problem. Several methods are used to assess whether crossflow exists
In addition to the injectivity/productivity calculations between strata, including (1) pressure tests between zones, (2)
discussed above, several other methods can be used to various logs for determining fluid saturations, permeability,
determine if fractures or fracture-like features are the source of porosity, and lithology, (3) injection/production profiles, (4)
the water problem. These other methods include (1) core and simulation, and (5) seismic methods. The most straightforward
log analyses (especially from highly deviated or horizontal method tests pressure differences between zones. Commonly,
wellbores), (2) pulse tests/pressure transient analyses, and (3) a packer is placed between two zones and one of the zones is
interwell tracer studies. allowed to pressure up. If a significant pressure can be
Various logging methods have been used to detect and maintained across the packer, effective barriers to crossflow
characterize fractures (Chapter 3 of Ref. 15). However, these exist between the zones. If a pressure difference cannot be
methods must be used with caution since they usually measure maintained, crossflow between the zones may occur. If the
properties at or near the wellbore. The value of these methods operator does not know whether crossflow occurs, he should
can be increased if the wellbore is deviated to cross the assume that crossflow exists.
different fracture systems (i.e., fractures with different
orientations). WOR History Plots. Plots of water/oil ratio (WOR) versus
Pressure transient analyses have often been used to time can provide a valuable indication of when an excess
characterize fractured reservoirs (Chapter 4 of Ref. 15). water problem develops.2,20 Along with other information,
such plots can also aid in identifying the cause of the problem.
SPE 70067 A STRATEGY FOR ATTACKING EXCESS WATER PRODUCTION 5

However, these “diagnostic plots” (of WOR or WOR wellbore pressure gradients, a small radius of penetration (e.g.,
derivative versus time) should not be used alone to diagnose ~1 ft) may be adequate to stop flow. Consequently, gelant
excessive water production mechanisms and problems.21,22 volumes can be quite small. Of course, greater gel volumes
This method was said to be capable of distinguishing whether and/or other treatment methods may be needed if flow behind
a production well is experiencing premature water pipe or fractures exist in the vicinity of the casing leak.
breakthrough caused by water coning or channeling through What placement and permeability reduction properties are
high permeability layers.20 According to this method, desired for gels used to plug casing leaks? Since the objective
gradually increasing WOR curves with negative derivative is to achieve total water shutoff from the leak and since small
slopes are unique for coning problems, and rapidly increasing gel volumes are often used for this application, the gel plug
WOR curves with positive derivative slopes are indicative of a should be relatively strong and must have a very low
channeling problem. As far as we are aware, this method has permeability. Rigid gels can be prepared from several
not been used to distinguish between linear flow (fracture or materials that yield permeabilities in the low microdarcy
flow behind pipe) and radial flow for either channeling or range.29,30 Gels for this application have often been formulated
coning. As mentioned above, the linear/radial distinction is with relatively high concentrations (4-7%) of acrylamide
extremely important—much more so than whether the polymers having a relatively low molecular weight (on the
problem is due to generic channeling or coning. order of 250,00 to 500,000 daltons).31 Gelants for this
Recently, reservoir models were built for water coning and application should be of relatively low viscosity and
channeling, respectively, and a sensitivity analysis was experience essentially no crosslinking of the polymer during
performed using numerical simulation.21,22 Reservoir and fluid gel treatment placement.
parameters were varied to examine WOR and WOR derivative
behavior for both coning and channeling production problems. Flow Behind Pipe (Problems 2 and 5 in Table 1). Problems
The results from this study demonstrated that multi-layer with unrestricted flow behind pipe are usually treated with
channeling problems could easily be mistaken as bottomwater cement.23 Cement can perform extremely well for this type of
coning, and vice versa, if WOR diagnostic plots are used alone application if the channel to be plugged is not too narrow (i.e.,
to identify an excessive water production mechanism. Hence, Problem 2). When narrow channels are encountered (Problem
WOR diagnostic plots can easily be misinterpreted and should 5, such as micro-annuli between cement and the formation or
therefore not be used alone to diagnose the specific cause of a the pipe), cement often cannot be placed effectively through
water production problem. small or constricted flow paths. Gels provide a better solution
for this case, since they can flow or extrude readily through
Solutions To Specific Types Of Problems narrow constrictions.32,33 The ability of gels to withstand high
After diagnosing the cause(s) of the excess water production, pressure gradients increases with decreasing channel width.34
what approach should be taken to solve the problem? As Therefore, gel alone cannot be expected to plug large voids
mentioned earlier, each problem type usually requires a behind pipe. In some cases, gelants or gels were injected first
different approach, including (1) choice of treatment method, (to penetrate into narrow constrictions), and cement was
(2) properties of the conformance or blocking agent, (3) injected subsequently to fill and plug larger near wellbore
volume of conformance or blocking agent used, and (4) voids and to prevent gel from washing out from their strategic
placement method. The remainder of this paper will focus on locations.35
the use of gelant or gel treatments, and will address whether When treating flow behind pipe problems where a
and how these treatments should be applied to successfully substantial drawdown pressure (i.e., >100 psi) exists, gelants
treat each of the problem types listed in Table 1. are often employed rather than preformed or partially formed
gels. Three reasons often favor gelant injection when treating
Casing Leaks (Problems 1 and 4 in Table 1). The most this problem type. First, flow constrictions in small flow
common methods to repair casing leaks (i.e., for Problem 1) channels behind pipe may prevent full penetration of
involve either cement23,24 or mechanical patches.2,25 However, preformed gel into the offending channels. These constrictions
these methods have generally not been very successful when do not significantly impede flow and placement of gelant
treating small casing leaks, such as “pinhole” or thread leaks solutions. Second, gelant invasion into permeable matrix rock
(Problem 4). In particular, cement has difficulty penetrating adjacent to the channel behind pipe is usually beneficial when
through small leaks. With luck, cement may lodge in and plug treating this type of problem. In contrast, preformed gels will
the leak, but small mechanical shocks often easily dislodge the not penetrate appreciably into the permeable matrix. Third,
cement plug. Gel treatments can be more successful for these because of relatively high near-wellbore drawdown pressures,
applications.26-28 Appropriately designed gelants flow easily gel in the channel probably will washout much more easily
through the small casing leaks and some distance into the than gel formed in the permeable matrix. Methods of sizing
formation surrounding the leak. Thus, the gel treatment is gel treatments for these applications have been strictly
directed at stopping flow in the porous rock around the empirical to date.
vicinity of the casing leak, rather than solely attempting to In certain circumstances, properly formulated gels of
permanently plug the casing leak itself. If the resultant gel preformed or partially formed crosslinked organic polymer
(placed in the matrix reservoir rock) can withstand the near
6 R.S. SERIGHT, R.H. LANE, AND R.D. SYDANSK SPE 70067

gels may be favored when treating long intervals of micro- technique that must be carefully designed and tailored to
annuli between the primary cement and the formation. individual well problems and often requires computer
simulation support for its successful implementation.
Unfractured Wells with Effective Barriers to Crossflow For gel applications in unfractured injection or production
(Problem 3). Often, when radial flow exists around a well wells where crossflow does not occur, how much gel should
(i.e., fractures are not important), impermeable barriers (e.g., be injected and what properties should the gel have? This
shale or anhydrite) separate hydrocarbon-bearing strata from a question is easily answered by considering Fig. 1, which was
zone that is responsible for excess water production. When the generated using the Darcy equation for radial flow.39 This
water zone is located at the bottom of the well, cement or sand figure applies to gel treatments both in injection and
plugs are used most commonly to stop water production. productions wells.
When the water zone is located above an oil zone, historically Fig. 1 plots the fraction of original injectivity or
the most common water shutoff methods include cement or productivity retained after a polymer or gel treatment as a
carbonate squeezes (into perforations) or mechanical packers function of the residual resistance factor (i.e., the permeability
or patches23—i.e., the conventional treatments of Category A. reduction provided by the polymer or gel). Fig. 1 applies to a
However, gels, involving gelant injection, have also been waterflooded reservoir with a 40-acre, 5-spot pattern with a
used frequently to treat these problems.7,28,32,36 In these unit-mobility displacement. The wellbore radius was 0.33 ft.
instances, the problem solution falls into Category B of Table Two cases of radii of gelant penetration (rgel) are presented5
1. Gels have two advantages over cements and carbonates for ft and 50 ft. A comparison of these two curves reveals that for
some applications.32 First, gelants can flow into porous rock, a given residual resistance factor, the injectivity or
whereas cements and particulate blocking agents are filtered productivity losses are not strongly dependent on the radius of
out at the rock surface. Cements (including “micro-fine gelant penetration. Therefore, the performance of the gelant
cement”) only invade and plug porous rock or sands of normal treatment is not sensitive to the volume of gelant injected. A
permeabilities (e.g., sandstone and sands of <1,000 mD) to five-foot radius of penetration will often be adequate for many
any significant distance by fracturing or parting the rock or applications, if the gel can withstand the high pressure
sand when sufficient injection pressures are provided. If the gradients near the wellbore. Fig. 1 also indicates the desired
cement does not adhere adequately to the rock in the properties of the gel. In the water zones, for the typical range
perforation or other large void (e.g., because of chemical of gelant penetrations, residual resistance factors of 20, 50,
incompatibility or mechanical shock), the zone may not be and 100, will provide water productivity losses of 80%, 90%,
sealed sufficiently. In contrast, gels (i.e., after gelation) can and 95%, respectively. These values are adequate for most
form an impermeable rubbery mass that extends past the rock radial flow problems.
surface, well into the porous rock. Second, gelants and gels In some cases where cold water is injected into wells in hot
can penetrate into and plug narrow channels (e.g., micro- reservoirs, thermal fractures may develop and extend a
annuli) behind pipe in the vicinity of the zone to be shut off.32 significant distance (e.g., 10 to 100 ft or more) from the
Therefore, in some cases, gels can provide a more effective wellbore.40,41 In these circumstances, the gel treatment should
seal in the zone to be plugged. plug both the matrix and the fractures in the offending zone.
When treating radial flow problems using gels or similar Many polymers and gels can reduce permeability to water
blocking agents, hydrocarbon zones must be protected during (kw) more than that to oil (ko) or gas (kgas). For the credible
gelant placement. Otherwise, the blocking agent will probably experimental data reported to date, polymers and gels may
also damage the hydrocarbon zones.14 Mechanical isolation of reduce kw more than ko, however, they always reduce ko to
zones is the most obvious method to protect oil zones during some extent. In the best cases, Zaitoun and Kohler42 reported
gelant placement. However, other methods exist—notably that adsorbed polymers significantly reduced kw at any given
dual injection.7,37,38 As an example of dual injection, gelant water saturation, while the oil relative-permeability curve was
might be injected down coiled tubing into the water zone basically unaffected by the polymer. However, the polymer
while non-damaging water or hydrocarbon fluid is injected increased the irreducible water saturation, thus lowering the
simultaneously down the annulus into the oil zone (while the endpoint relative permeability to oil. Therefore, for all
two zones are in fluid communication). Downhole pressure practical purposes in zones with high oil saturations, the
gauges in the tubing and annulus are carefully monitored to polymer treatment reduces the effective permeability to oil to
maintain a very delicate pressure balance. Near the wellbore, some extent.
this balance minimizes gelant crossflow into the oil zones and For gel treatments applied to water injection wells, the
protective-fluid crossflow into the water zone. This method is disproportionate permeability reduction is of no value.
of particular interest and value for wells where mechanical However, in production wells, the property is critical to the
zone isolation is impractical, especially gravel-packed wells success of gel treatments if hydrocarbon zones are not
and wells with flow behind pipe. The method and its protected during gelant placement. Even then, the property is
associated gel treatment will not be effective in cases where of value only when zones with high hydrocarbon saturation
laterally extensive barriers (e.g., shale or anhydrite layers) are are distinct from the offending water producing zones.43 In
not present out away from the wellbore.19 The dual injection other words, this “disproportionate permeability reduction”
technique is considered to be an advanced zone isolation
SPE 70067 A STRATEGY FOR ATTACKING EXCESS WATER PRODUCTION 7

will not mitigate water production from a reservoir that has When treating water production problems in unfractured
effectively only one zone. When a single zone exists, even if reservoirs with barriers to crossflow, gel treatments can be
the polymer or gel can significantly reduce permeability to applied in either injection or production wells.
water without affecting the permeability to oil, the average
fractional flow of water and oil from that zone must remain 2-D Coning: Hydraulically Fractured Production Wells
the same. If the polymer or gel near a production well allows (Problem 6). When production wells are hydraulic fractured,
oil to pass but not water, the water saturation will increase the fracture often unintentionally breaks into water zones,
near and just beyond the gel bank, thus, decreasing the relative causing substantially increased water production. Gelant
permeability to oil until the fractional water and oil flows treatments have significant potential to correct this problem.
match the values that existed before the polymer or gel These gelant treatments rely on the ability of these gels to be
treatment. Therefore, unless a particular zone is at its placed in the rock matrix adjacent to the fractures and to
irreducible water saturation, a polymer or gel treatment will reduce permeability to water much more than that to
always cause some loss of oil productivity, even if the hydrocarbon (disproportionate permeability reduction). An
polymer or gel reduces kw without affecting ko. This loss of oil engineering-based method was developed for designing and
productivity necessarily will be in direct proportion to the loss sizing gelant treatments in hydraulically fractured production
of water productivity caused in that particular zone. wells.45 This design procedure was incorporated in user-
A common misconception is that the disproportionate friendly graphical-user-interface software that can be
permeability reduction will be of value mainly in treating downloaded from the internet at http://baervan.nmt.edu/randy.
unfractured production wells where fluid flow is radial around In these matrix rock treatments, gelants flow along the
the wellbore. However, two technical obstacles currently fracture and leak off a short, predictable distance into the
impede this type of treatment from being commonly matrix rock of all the zones (water, oil, gas). Success for such a
successful. First, if zones are not isolated during gelant treatment requires that the gel reduce permeability to water
placement, then generally, the residual resistance factor much more than that to hydrocarbon in the treated matrix rock.
(permeability reduction value) in the oil zone must be less The ability of the gel to stop water entry into the fracture is
than two while the residual resistance factor in the water zone determined by the product of gelant leakoff distance (from the
must be greater than 10. The reason for this requirement can fracture face) and the residual resistance factor (permeability
be appreciated by considering Fig. 1. For radial flow, reduction factor) provided by the gel. For example, consider the
relatively small residual resistance factors (Frr) can cause case where the gelant leaks off 0.2 ft into both water and oil
significant injectivity or productivity losses. For example, for zones, and in the gel-contacted rock, permeabilities to water and
a gel radius of 50 ft, a Frr value of 2 causes a 27% loss in oil are reduced by factors 50,000 and 50, respectively. (These
productivity, while a Frr value of 10 causes a 75% loss. Both properties have been reported for a gel formulation.46) In this
of these losses might be considered unacceptable if these are case, the gel only adds the equivalent of 10 feet of additional
oil zones. Thus, in unfractured wells, oil residual resistance rock that the oil must flow through to enter the fracture (i.e.,
factors (Frro) provided by the gel must be small. 0.2 ft x 50). In contrast, for the water zone, the water must
A second technical obstacle also thwarts the flow through the equivalent of 10,000 ft of additional rock to
disproportionate permeability reduction from being usable in enter the fracture (i.e., 0.2 ft x 50,000). Thus, in this
practice when treating radial flow problems. Especially for circumstance, the gel can substantially reduce water
gels and/or products of gelation reactions, Frro values less than production without significantly affecting oil productivity.
two may be difficult to achieve in a predictable and In this method, fluid entry into the fracture is controlled by
controllable manner.29,44 Low Frro values usually mean that gel in rock next to the fracture.45 Ideally, fracture conductivity
gelation was incomplete and the products of the gelation is not reduced significantly, since it allows a conductive path
reaction were small gel particles that become trapped in pore for oil flow into the wellbore. To some extent, gravity
throats. These particles occupy a small fraction of the aqueous segregation of the gelant (between placement and gelation)
pore space. Gelation reactions are usually sensitive to pH, will mitigate damage to the fracture when the excessive water
salinity, and other factors, and these factors are influenced by production originates from an underlying aquifer. However, to
the rock lithology and resident fluid composition.29,44 minimize fracture damage, an oil or water post-flush could be
Consequently, small differences in rock lithology and used to displace gelant from the fracture.
reservoir conditions may significantly change the From a rigorous viewpoint, the method assumes that
concentration and size of particles formed during the early impermeable barriers (e.g., shale or calcite) separate adjacent
stages of gelation—ultimately resulting in residual resistance zones.45 However, the method should frequently provide
factors that are unpredictable and uncontrollable. acceptable predictions even if crossflow can occur between the
As will be discussed in the next section, the disproportionate water bearing and oil bearing zones. For example, consider the
permeability reduction currently is of much greater value in case where oil lies on top of water in a single formation (i.e., a
treating linear flow problems (i.e., fractured production wells) common situation where coning becomes a problem). Previous
than radial flow problems. work43,46 showed that gravity alone can retard water influx into
oil zones much more effectively when the water must “cusp” to
a linear pressure sink (i.e., a vertical fracture or a horizontal
8 R.S. SERIGHT, R.H. LANE, AND R.D. SYDANSK SPE 70067

well) than when the water “cones” to a point pressure sink (i.e., Phillips applied 37 treatments in Arbuckle formations using
a partially penetrating vertical well). For the type of gel eight different organic polymer and polymer-crosslinker
treatment that we are proposing for application in hydraulic combinations.47 In their treatments, the average incremental
fractures, in many cases, gravity may be sufficient to minimize recovery was 1.9 STB/lb polymer, with a range from -1 to 13
water invasion into the hydrocarbon zones of a single formation. STB/lb. The average time for the well to return to the pre-
Of course, the degree of water invasion (coning) into treatment WOR, and oil production rate was 12 months, with a
hydrocarbon zones increases with increased production rate, range from 2 to 43 months. The treatments typically reduced
pressure drawdown, vertical formation permeability, and total fluid productivity by a factor of two. Interestingly,
hydrocarbon viscosity, and decreases with increased water- Phillips found that the incremental oil recovery, treatment
hydrocarbon density difference, horizontal formation lifetime, and WOR reduction did not correlate with the mass
permeability, and oil column thickness.43,46 If water invades too of polymer injected (390 to 1,400 lbs/well), type of polymer or
far into the hydrocarbon zone, a water block could form that gel treatment (8 types used), productivity reduction induced by
reduces hydrocarbon productivity. the treatment (1 to 5), structural position of the completion,
To use this procedure to reduce water production from a completion type, fluid level before the treatment, or the
hydraulic fracture, field data are needed, coupled with results Arbuckle reservoir.47 (Treatments were applied in several
from two simple laboratory experiments.45 The needed field Arbuckle reservoirs.)
data include: (1) fluid production rates before the gel A review of 274 water shutoff treatments that were applied
treatment, (2) downhole static and flowing pressures before between 1970 and 1990 focused on gel treatments in two
the gel treatment, (3) permeabilities, porosities, and thickness naturally fractured carbonate formations (Arbuckle and
of the relevant zones, (4) water and oil viscosities at reservoir Ellenberger).4 For the results published, the median WOR was
temperature, and (5) well spacing or distance between wells. 82 before gel injection, 7 shortly after gel treatment, and 20
These parameters are often available during conventional gel after one to two years following the treatment. The median oil
treatments. The downhole pressure drops are critically productivity increased by 3 shortly after treatment and
important for this method. They must be reasonably current returned to pre-treatment levels after one to two years.
and measured specifically for the well to be treated. The positive effects of these treatments were generally
Use of the procedure also requires oil and water residual short-lived in the Arbuckle and Ellenberger formation.
resistance factors from laboratory core experiments.45 These However, for several gel applications in the Madison
experiments must be conducted using the gelant, oil, brine, formation in Wyoming, reductions in water cut were sustained
rock, and temperature that are representative of the intended for many years.46 Chromium(III)-carboxylate/acrylamide-
application. In the absence of laboratory oil and water residual polymer gel water shutoff treatments also were applied to 14
resistance factors, the model can use field data to back- economically marginal production wells of the old and mature
calculate these values in situ after a gel treatment. This Big Lake field in Texas.9 Water production was decreased, on
information may be useful when designing similar treatments the average, from 3,410 to 993 BWPD and oil production was
in nearby wells. These calculations have also been increased, on the average, from 2 to 14 BOPD. The main
incorporated into the software. For cases where residual producing zone of these 14 oil wells was the dolomitic
resistance factors are calculated from field data, three Grayburg formation that was naturally fractured. Excess water
parameters (from a similar, previous gelant treatment) are production was believed to be coning up through vertical
required in addition to the five items listed in the previous fractures from the underlying active aquifer.9,63 During these
paragraph. These three parameters are (1) fluid production successful gel treatments applied to the Madison and
rates after the gel treatment, (2) accurate downhole static and Grayburg formations, partially formed gels were injected.
flowing pressures after the gel treatment, and (3) the volume Thus, the gel solution to these two excess water production
of gelant injected. problems (Problem 7) shifts into Category C of Table 1.
Although somewhat challenging to properly design and Results from treatments applied to Problem 7 raise a number
execute, strong and/or rigid gel treatments, involving the of important questions. First, what is the water shutoff
injection of partially formed gels, can be used to treat 2-D mechanism for these treatments? Do the treatments work
water coning in hydraulically fractured production wells. In primarily because gelant penetrates into the porous rock and
this treatment strategy, gravity is exploited to selectively place provides disproportionate permeability reduction? Or do the
a partially gelled solution in the lower portion of the fracture.9 treatments work because gels selectively plug the lower parts
of the fracture system more than the upper parts? Is it better to
Natural Fracture System Leading to an Aquifer (Problem inject a gelant that forms a strong gel or a weak gel? Why
7). Several operators reported impressive (but often short- were the benefits from the treatments temporary in most
lived) results from polymer and gel treatments in production cases? How should these treatments be sized? Should
wells in the Arbuckle, Ellenberger, and Madison preformed gels be injected instead of gelants? Unfortunately,
formations.4,47,48 These treatments were applied to reduce these questions remain unanswered for the present.
excessive water production emanating via natural fractures
from underlying aquifers that provided strong water drives. Individual Fractures that Cause Channelling from
Injectors to Producers (Problem 9). Gel treatments currently
SPE 70067 A STRATEGY FOR ATTACKING EXCESS WATER PRODUCTION 9

provide the most effective means to reduce channeling fracture width and gel properties (i.e., gel composition and
through fractures.4,49-51 Except in narrow fractures (i.e., rigidity). For a typical Cr(III)-acetate-HPAM gel (containing
fracture widths less than 0.02 in.), extruded gels have a 0.5% polymer), a 2 psi/ft pressure gradient was noted during
placement advantage over conventional gelant treatments extrusion through a 0.1-in.-wide fracture.34 Therefore, in field
when treating channeling through fractures. To explain, during applications, knowledge and/or estimation of fracture widths
conventional gel treatments, a fluid gelant solution typically is important for deciding the composition and properties of the
flows into a reservoir through both the porous rock and the gel to be injected.
fractures. After placement, chemical reactions (i.e., gelation) For interwell channeling, the effective average width of the
cause an immobile gel to form. During gelant injection, fluid most direct fracture can be estimated from interwell tracer
velocities in the fracture are usually large enough that viscous tests.54,55 Tester et al.54 suggested that the best estimate of the
forces dominate over gravity forces.52 Consequently, for volume of a fracture path is provided by the modal volume
small-volume treatments, the gelant front is not greatly (i.e., the volume associated with the peak concentration in the
distorted by gravity during gelant injection. However, after produced tracer distribution). The interwell tracer time (t in
gelant injection stops, a small density difference (e.g., 1%) days) associated with this peak concentration can be use to
between the gelant and the displaced reservoir fluids allows estimate effective average fracture width (wf in inches)55:
gravity to rapidly drain gelant from at least the upper part of
the fracture.52 Generally, gelation times cannot be controlled wf = 5.4 x10-5 Lf [µ /(t∆p)]1/2 ,............................................ (4)
well enough to prevent gravity segregation in the time
between gelant injection and gelation. where Lf is the injector-producer well separation (in feet), µ is
Alternative to conventional gelant treatments, formed tracer fluid viscosity (in cp), and ∆p is the downhole interwell
(preformed) gels can be extruded through fractures. Since pressure drop (in psi).
these gels are 103 to 106 times mores viscous than gelants, For some applications where wide fractures or large vugs
gravity segregation for gels is much less important than for are present, gels alone may not provide sufficient mechanical
gelants. For some of the most successful treatments in strength and flow resistance to plug the channel. In these
fractured reservoirs, formed gels were extruded through cases, particulate matter (sand, cellophane, fibers, nut shells,
fractures during most of the placement process.11,49-51 etc.) can be added to increase the mechanical strength and
The extrusion properties of a Cr(III)-acetate-HPAM plugging characteristics of the gel.56-58
(chromium(III)-carboxylate/acrylamide-polymer) gel have Gel jobs to treat individual fractures that cause channeling
been characterized as a function of injection rate and time and from injectors to producers can be applied in either injection
fracture width and length.34 Gels concentrate or dehydrate or production wells.
during extrusion through fractures. During flow in a fracture,
the rate of dehydration of these gels varies inversely with the Faults or Fractures that Cross Deviated or Horizontal
square root of time. This fact allows gel propagation along Wells (Problem 8). Deviated and horizontal wells are prone
fractures to be predicted.34,53 (See Figs. 2 and 3 for to intersect faults or fractures. If these faults or fractures
propagation of a Cr(III)-acetate-HPAM gel in a vertical connect to an aquifer, water production can jeopardize the
fracture of fixed height.) To maximize gel penetration along well.51 Often, the completions of these wells severely limits
fractures, the highest practical injection rate should be used. the use of mechanical methods to control fluid entry. In
However, in wide fractures or near the end of gel injection, gel contrast, gel treatments can provide a viable solution to this
dehydration may be desirable to form stronger and rigid gels type of problem. However, conventional gelant treatments are
that are less likely to washout after placement. In these not the desired form of remediation in this case. In a
applications, reduced injection rates may be appropriate. In conventional gelant treatment, a fluid gelant solution is
single, wide (i.e., >0.5 in.) vertical fractures (of fixed height) injected that flows down the well into the target fracture or
where short distances of penetration are needed, the gel fault and also leaks off into the porous rock around the
volume required increases roughly with the distance of wellbore and the fracture or fault. The resultant gel may plug
penetration. In single vertical fractures (of fixed height) with or severely restrict water entry into the fracture or fault.
narrow to moderate widths (i.e. 0.02 to 0.5 in.), the required Unfortunately, the gelant will also flow into the exposed
gel volume increases roughly with the distance of penetration hydrocarbon bearing rock all along the well during the
raised to the 1.5 power. A rule of thumb derived from this placement process. Consequently after gelation, oil
latter behavior is that doubling the distance of penetration productivity can be damaged as much as water productivity.
along a given fracture (of narrow or moderate width) requires Alternatively, a formed gel can be pumped down the well and
tripling the volume of injected gel. selectively placed in the fracture.34,51,53 The gel formulation
A minimum pressure gradient is required to extrude a given may exist as an uncrosslinked fluid at the wellhead, so long as
gel through a fracture.34 After this minimum pressure gradient significant gelation occurs before the gelant reaches the oil
is met, the pressure gradient during gel extrusion is insensitive zone. Then, because formed gels do not enter or flow through
to the flow rate. The pressure gradient required for gel porous rock,59 damage to oil productivity can be minimized. In
extrusion varies inversely with the square of fracture width.34 contrast, the gel can extrude selectively into and plug the
The volume of gel that can be injected depends critically on
10 R.S. SERIGHT, R.H. LANE, AND R.D. SYDANSK SPE 70067

fracture or fault. When the well is returned to production, gel gel injection, and (dp/dl)gel is the pressure gradient required for
remaining in the wellbore can often flow back to the surface. gel extrusion through the fracture of interest. As mentioned
If designed properly, gel in the fault or fracture will remain in earlier, the pressure gradient for gel extrusion varies inversely
place because the fracture width is much smaller than the with the square of fracture width.34 For one Cr(III)-acetate-
diameter of the wellbore. (The pressure gradient required to HPAM gel (with 0.5% HPAM) that is commonly used in field
mobilize formed gels varies inversely with the square of applications, the pressure gradient (in psi/ft) for gel extrusion
fracture width or tube diameter.34) Alternatively, coiled tubing is related to fracture width (in inches) using Eq. 6.
can be used to circulate gel out from the wellbore.32 (In
practice, water, oil, or an uncrosslinked polymer solution is (dp/dl)gel = 0.02 / (wf )2 ...................................................... (6)
often injected immediately after the gel in an attempt to
displace gel from the wellbore into the fracture.51 Since this Of course, the coefficient in Eq. 6 (e.g., 0.02) depends on gel
displacement is unstable, its effectiveness can be questioned.) composition. More rigid gels exhibit greater coefficients and
If the water production problem is caused by a single pressure gradients during extrusion.
fracture or fault that intersects the horizontal wellbore, the
distance of gel penetration into the fracture need not be Injector-Producer Channeling in Naturally Fractured
particularly large.60 In this case, the benefit gained varies Reservoirs (Problem 10). Some of the most successful gel
approximately logarithmically with the distance of gel treatments were applied to reduce water and gas channeling in
penetration.53 However, this conclusion is specific to one naturally fractured reservoirs.11,49,50,62,63 The primary objective
particular scenario—i.e., a single fault or fracture intersecting of these gel treatments was to improve sweep efficiency and to
a horizontal well. The conclusion may not be valid for vertical promote incremental oil production. A secondary benefit of
wells or if multiple fractures or faults intersect a horizontal the gel treatments was the substantial reduction of excessive
well, or if a natural fracture system is present. Furthermore, water and gas production at the offsetting production wells.
even for the case of a single fault or fracture that intersects a During these injection well applications, the time required to
horizontal well, some value may be realized by injecting a inject large volumes (e.g., 10,000 to 37,000 bbls) of gel was
significant amount of gel to mitigate the possibility of gel typically greater than the gelation time by a factor around
washout after the well is returned to production. 100.11,49,50 Thus, formed gels extruded through fractures
For horizontal wells that cross faults or fractures, simple during most of the placement process. Several operators
calculations based on productivity data can give at least a reported that oil recovery increased with increased volume of
rudimentary indication of the width of the fracture that causes gel injected per treatment.11,49,50 However, sizing of these
the excess water production.53 The calculations can also give treatments to date has been empirical—dictated primarily by
an idea of how far the gel should penetrate to provide a perceived economic and operational limitations. Engineering-
beneficial effect.34 Using laboratory data coupled with field based sizing methods are under development for this type of
data collected before, during, and after gel injection of similar problem.55
gel treatments, the calculations can also give an indication of Theoretical work indicates that gel treatments have the
how far the gel actually penetrated into the fracture.53 To greatest potential when the conductivities of fractures that are
successfully make these determinations, accurate flowing and aligned with direct flow between an injector-producer pair are
static downhole pressures are critical measurements that must at least 10 times the conductivity of off-trend fractures.55 Gel
be obtained during field applications of these gel treatments. treatments also have their greatest potential in reservoirs with
In vertical fractures that cut through vertical wells, gel flow moderate to large fracture spacing. Produced tracer
in the fracture is generally linear. However, in vertical concentrations from interwell tracer studies can help identify
fractures that cut through horizontal wells, the flow geometry reservoirs that are good candidates for water shutoff using gel
is radial (at least, near the well). During gel extrusion through treatments. The average width of the most direct fracture
fractures of a given width, the pressure gradient and degree of between an injector-producer pair can be estimated from the
gel dehydration were nearly independent of position and breakthrough time from an interwell tracer study using Eq. 4.
velocity during both radial and linear flow.61 Because the Since the ability of a gel to extrude through a fracture depends
pressure gradient during gel extrusion is almost independent critically on the fracture width or conductivity,34,59,61 this
of injection flux, the pressure gradient is nearly independent of knowledge is important when selecting an appropriate gel
radial position from the wellbore. Thus, the distance of gel formulation for the treatment.
penetration from the wellbore (Lgel or rgel) can be estimated Simulation studies indicate that the potential for successful
regardless of whether flow in the fracture is linear or radial. application of a gel treatment becomes greater as the peak
produced tracer concentration increases above 20% of the
Lgel or rgel = (∆pgel - ∆pwater) (dp/dl)gel ,...............................(5) concentration the injected tracer concentration.55 When
produced tracer concentrations are low (i.e., less than 1% of
where ∆pwater is the pressure drawdown (i.e., the downhole the injected tracer concentration), gel treatments are unlikely
pressure difference between the wellbore and the formation) to be effective. However, results from a poorly designed tracer
during water injection, ∆pgel is the pressure drawdown during test can mislead one to believe that a gel treatment has little
potential. For example, if the tracer bank is too small,
SPE 70067 A STRATEGY FOR ATTACKING EXCESS WATER PRODUCTION 11

dispersion can reduce produced tracer concentrations to very engineering calculations indicate a very low probability for
low values in a fracture system even though a gel treatment success for gel treatments in three-dimensional coning
has excellent potential. applications? The answer may be tied to two special
Gel treatments to reduce injector-producer channeling in characteristics of this field. First, Shell’s simulation work
naturally fractured reservoirs can be applied in either injection suggests that effective barriers to vertical flow are present.65
or production wells. These barriers were not recognized when the first treatments
were applied. Second, the oil viscosity was about 80 cp. Thus,
Three-Dimensional Coning and Cusping (Problems 11 and viscous fingers of water may have arrived at a given well
12). Gelant or gel treatments have an extremely low much earlier in some of the discrete zones than others.
probability of success when applied toward cusping or three- Because the oil was much more viscous than the gelant (~10
dimensional coning problems occurring in unfractured matrix cp), the gelant may have followed these water fingers and
reservoir rock. When treating coning problems, a common preferentially reduced flow in the water zones to a much
misconception is that the gelant will only enter the water greater extent than if a light oil was present. This scenario is
zones at the bottom of the well. In reality, this situation will consistent with basic reservoir engineering calculations.14,43 Of
occur only if the oil is extremely viscous and/or the aqueous course, this scenario suggests that the real problem in this
gelant is injected at an extremely low rate (to exploit gravity reservoir was not three-dimensional coning, but rather viscous
during gelant placement). In the majority of field applications fingering through discrete high permeability pathways. Thus,
to date, the crude oils were not particularly viscous, and gelant consistent with our original contention, gelant treatments are
injection rates were relatively high. Consequently, one must not likely to be effective against three-dimensional coning.
be concerned about damage that polymer or gel treatments Gel treatments are also expected to be ineffective when
cause to hydrocarbon-productive zones. treating cusping. In cusping, like three-dimensional coning,
Even if a polymer or gel reduces kw without affecting ko, gel the well is produced so rapidly that viscous forces overcome
treatments have limited utility in treating 3-D coning gravity forces. For cusping in particular, water from an aquifer
problems. Extensive numerical studies using a variety of follows an inclined zone up to the well. The only practical
coning models indicate that gel treatments can only provide method to stop water production from the zone (other than
improvement if the desired production rate is less than 1.5 to 5 decreasing the production rate) is to plug the zone. Unless
times the pretreatment critical rate.43,65 This circumstance extraordinary circumstances exist (as in the Marmul case
rarely occurs. above), hydrocarbon-productive zones in radial flow must be
In contrast to the very limited potential of polymers and gels protected during gelant placement. (For the Marmul
in successfully treating 3-D coning, these treatments have treatments, one wonders whether the success rate might have
much greater potential for successfully treating “two- been 14/14 instead of 5/14 if hydrocarbon zones had been
dimensional coning” where vertical fractures cause water from protected during gelant placement.)
an underlying aquifer to be sucked up into a well. Whereas gel
treatments will only raise the critical rate by factors from 1.5 Injector-Producer Channeling in Unfractured Reservoirs
to 5 in unfractured wells, they can raise the critical rate by with Crossflow (Problem 13). Gelant and gel treatments are
more than a factor of 100 in fractured wells.43,65 expected to be ineffective for treating injector-producer
A number of literature reports suggested that gel or foam channeling in unfractured reservoirs where fluids can
treatments were effective in mitigating 3-D coning. A critical crossflow between zones.19 For many years, engineers
examination of these reports64 revealed that they fall into one recognized that near wellbore blocking agents are ineffective
of three categories: (1) evidence suggests that flow behind in these applications.66 Even if the blocking agent could be
pipe or fractures or fracture-like features were the actual cause confined only to the high permeability channel, water quickly
of the “coning,” (2) results were not convincing that the cross flows around any relatively small plug. The only hope
treatment reduced the water/oil ratio, gas/oil ratio, or for blocking agents in these applications exists if a very large
water/gas ratio, or (3) insufficient evidence was presented to plug (i.e., that plugs most of the channel) can be selectively
determine whether the problem was caused by three- placed only in the high permeability zone.66 Unfortunately,
dimensional coning, flow behind pipe, or flow through existing gelants (including the so-called “colloidal dispersion
fractures or fracture-like features. gels”) enter and damage all open zones in accordance with the
Shell’s (PDO) experience in the Marmul field provides an Darcy equation and basic reservoir engineering principles.19
interesting exception to the above observations.65 Five of Penetration and damage caused to the less-permeable zones is
fourteen gel treatments were quite successful in reducing the greater for viscous gelants than for low-viscosity fluids. Also,
water cut—up to 45% in one case. Convincing evidence was penetration and damage caused to the less-permeable zones is
presented that flow behind pipe and fracture-like features were greater when crossflow can occur than when crossflow cannot
not important. Gelant (0.4% to 0.5% cationic polyacrylamide occur.19 Although an admirable attempt was made to devise a
with glyoxal as a crosslinker) was bullheaded into the wells, sophisticated process where gelant treatments might be
using 700 to 2,500 bbls per treatment (11 to 19 bbls per ft in effective in treating this type of problem,67,68 traditional
gravel packed completions). The key question is, why were polymer floods provide a more cost-effective and reliable
five of the treatments successful, when basic reservoir solution.19,69-71
12 R.S. SERIGHT, R.H. LANE, AND R.D. SYDANSK SPE 70067

Conclusions
1. When addressing excess water production problems, the References
easiest problems should be attacked first, and diagnosis of 1. “Profile of the Oil and Gas Extraction Industry,” EPA/310-R-
water production problems should begin with information 99-006 (Oct. 1999) 38.
already at hand. To facilitate implementation of this 2. Bailey et al.: “Water Control,” Oilfield Review 12 (Spring 2000)
strategy, a prioritization of water production problems was 30-51.
3. Elphick, J. and Seright, R.S.: “A Classification of Water
provided (Table 1). Problem Types,” Paper 1 presented at the 1997 PNC 3rd
2. Conventional methods (e.g., cement, mechanical devices) International Conference on Reservoir Conformance, Profile
normally should be applied first to treat the easiest Control, Water and Gas Shut Off, Houston, Aug. 6.
problems—i.e., casing leaks and flow behind pipe where 4. Seright, R.S. and Liang, J.: “A Survey of Field Applications of
cement can be placed effectively and unfractured wells Gel Treatments for Water Shutoff,” paper SPE 26991 presented
where flow barriers separate water and hydrocarbon zones. at the 1994 SPE III Latin American & Caribbean Petroleum
3. Gelant treatments normally are the best option for casing Engineering Conference, Buenos Aires, Argentina, April 27-29.
leaks and flow behind pipe with flow restrictions that 5. Chou, S.I. et al.: “Development of Optimal Water Control
prevent effective cement placement. Strategies,” paper SPE 28571 presented at the 1994 SPE Annual
Technical Conference and Exhibition, New Orleans, Sep. 25-28.
4. Both gelants and preformed gels have been successfully 6. Pappas, J., Creel, P., and Crook, R.: “Problem Identification and
applied to treat hydraulic or natural fractures that connect Solution Method for Water Flow Problems,” paper SPE 35249
to an aquifer. presented at the 1996 Permian Basin Oil & Gas Recovery
5. Treatments with preformed or partially formed gels Conference, Midland, March 27-29.
normally are the best option for faults or fractures crossing 7. Fulleylove, R.J. et al.: “Water Shut-Off in Oil Production
a deviated or horizontal well, for a single fracture causing Wells—Lessons from 12 Treatments,” paper SPE 36211
channeling between wells, or for a natural fracture system presented at the 1996 International Petroleum Exhibition and
that allows channeling between wells. Conference, Abu Dhabi, Oct. 13-16.
6. Gel treatments should not be used to treat the most difficult 8. Lane, R.H.: “Field Operational and Performance Issues of
Polymeric Water Control Agents,” paper SPE 37243 presented
problems—i.e., three-dimensional coning, cusping, or at the 1997 SPE International Symposium on Oilfield
channeling through strata with crossflow. Chemistry, Houston, Feb. 18-21.
9. Sydansk, R.D. and Southwell, G.P.: “More Than 12 Years’
Nomenclature Experience With a Successful Conformance-Control Polymer-
Frr = residual resistance factor Gel Technology,” SPEPF 15 (4) (Nov. 2000) 270-278.
h = height, ft [m] 10. Love, T. et al.: “Problem Diagnosis, Treatment Design, and
k = permeability, darcys [µm2] Implementation Process Improves Waterflood Conformance,”
paper SPE 49201 presented at the 1998 SPE Annual Technical
kgas = permeability to gas, darcys [µm2]
Conference and Exhibition, New Orleans, Sep. 27-30.
ko = permeability to oil, darcys [µm2] 11. Hild, G.P. and Wackowski, R.K.: “Reservoir Polymer Gel
kw = permeability to water, darcys [µm2] Treatments To Improve Miscible CO2 Flood,” SPEREE (April.
L = distance along a fracture, ft [m] 1999) 196-204.
Lgel = distance of gel penetration along a fracture, ft [m] 12. Mennella, A. et al.: “Candidate and Chemical Selection Rules
Lf = fracture length, ft [m] for Water Shutoff Polymer Treatments,” paper SPE 54736
∆p = pressure drop, psi [Pa] presented at the 1999 SPE European Formation Damage
Conference, The Hague, May 31-June 1.
∆pgel = pressure drop during gel injection, psi [Pa] 13. Soliman, M.Y. et al.: “Integration of Technology Supports
∆pwater = pressure drop during water injection, psi [Pa] Preventive Conformance Reservoir Techniques,” paper SPE
dp/dl = pressure gradient, psi/ft [Pa/m] 62553 presented at the 2000 SPE/AAPG Western Regional
q = total injection or production rate, BPD [m3/d] Meeting, Long Beach, June 19-23.
re = external drainage radius, ft [m] 14. Seright, R.S.: “Placement of Gels to Modify Injection Profiles,”
rgel = radius of gel penetration, ft [m] paper SPE/DOE 17332 presented at the 1988 SPE/DOE
rw = wellbore radius, ft [m] Enhanced Oil Recovery Symposium, Tulsa, April 17-20.
Sor = residual oil saturation 15. Aguilar, R.: Naturally Fractured Reservoirs, Pennwell, Tulsa
(1980).
t = time, days
16. Wagner, O.R.: “The Use of Tracers in Diagnosing Interwell
wf = fracture width, in. [m] Reservoir HeterogeneitiesField Results,” JPT (Nov. 1977)
µ = viscosity, cp [mPa-s] 1410-1416.
17. Datta-Gupta, A., Vasco, D.W., and Long, J.C.S.: “Sensitivity
Acknowledgments and Spacial Resolution of Transient Pressure and Tracer Data
Financial support for this work is gratefully acknowledged for Heterogeneity Characterization,” paper SPE 30589 presented
from the National Petroleum Technology Office of the United at the 1995 SPE Annual Technical Conference and Exhibition,
States Department of Energy, BP-Amoco, Chevron, China Dallas, Oct. 22-25.
National Petroleum Corp., Chinese Petroleum Corp., 18. Smith, L.R., Fast, C.R., and Wagner, O.R.: “Development and
Field Testing of Large Volume Remedial Treatments for Gross
Marathon, Shell, and Texaco.
Water Channeling,” JPT (Aug. 1969) 1015-1025.
SPE 70067 A STRATEGY FOR ATTACKING EXCESS WATER PRODUCTION 13

19. Sorbie, K.S. and Seright, R.S.: “Gel Placement in 38. Bergem, J. et al.: “Successful Water Shutoff in a High-
Heterogeneous Systems with Crossflow,” paper SPE 24192 Temperature, High-Volume Producer—A Case History from the
presented at the 1992 SPE/DOE Symposium on Enhanced Oil Ula Field, Offshore Norway,” paper SPE 38833 presented at the
Recovery, Tulsa, April 22-24. 1997 Annual Technical Conference and Exhibition, San
20. Chan, K.S.: “Water Control Diagnostic Plots,” paper SPE 30775 Antonio, Oct. 5-8.
presented at the 1995 SPE Annual Technical Conference and 39. Seright, R.S.: “Improved Methods for Water Shutoff,” DOE
Exhibition, Dallas, Oct. 22-25. Report DOE/PC/91008-1, Contract No. DE-AC22-94PC91008,
21. Seldal, M.: “Using Water/Oil Ratios to Diagnose Excessive BDM-Oklahoma Subcontract No. G4S60330, U.S. Department
Water Production Mechanisms,” Master of Science Thesis, New of Energy (Aug. 1997) 8-10.
Mexico Institute of Mining and Technology (May 1997). 40. Martins, J. P. et al: “Produced Water Reinjection and Fracturing
22. Seright, R.S.: “Improved Methods for Water Shutoff,” Annual in Prudhoe Bay,” SPERE (Aug. 1995) 176-182.
Technical Progress Report (U.S. DOE Report DOE/PC/91008- 41. Paige, R. W. and Murray, L. R.: “Re-Injection of Produced
4), U.S. DOE Contract DE-AC22-94PC91008, BDM-Oklahoma Water – Field Experience and Current Understanding,” Eurock
Subcontract G4S60330 (Nov. 1997). ‘94 (1994) Delft, Netherlands 731-738.
23. Marca, C.: “Remedial Cementing” in Well Cementing, 42. Zaitoun, A., Kohler N., and Guerrini, Y.: “Improved
Developments in Petroleum Science 28 (1990), E.B. Nelson, ed., Polyacrylamide Treatments for Water Control in Producing
Elsevier, Amsterdam, 13-1. Wells,” JPT (July 1991) 862-867.
24. Meek, J.W. and Harris, K.L.: “Repairing Casing Leaks Using 43. Liang, J., Lee, R.L., and Seright, R.S.: “Placement of Gels in
Small-Particle-Size Cement,” paper IADC/SPE 21972 presented Production Wells,” SPEPF (Nov. 1993) 276-284; Trans. AIME
at the 1991 IADC/SPE Drilling Conference, Amsterdam, March 295.
11-14. 44. Seright, R.S. and Martin, F.D.: “Impact of Gelation pH, Rock
25. Macrae, J.: “Zero Extrusion Casing Patch,” paper SPE 37650 Permeability, and Lithology on the Performance of a Monomer-
presented at the 1997 IADC/SPE Drilling Conference, Based Gel,” SPERE (Feb. 1993) 43-50.
Amsterdam, March 4-6. 45. Seright, R.S., Liang, J., and Seldal, M.: “Sizing Gelant
26. Urdahl, H. et al.: “Experience with Temporarily Sealing Treatments in Hydraulically Fractured Production Wells,”
Leaking Tubing Annuli With Extended-Life Polymer Gel Plugs SPEPF (Nov. 1998) 223-229.
in the Greater Ekofisk Area,” paper SPE 24982 presented at the 46. Seright, R.S., Liang, J., and Sun, H.: “Gel Treatments in
1992 European Petroleum Conference, Cannes, France, Nov. Production Wells with Water Coning Problems,” In Situ (1993)
16-18. 17(3), 243-272.
27. Creel, P. and Crook, R.: “Gels, Monomer Solutions Fix Pinhole 47. Moffitt, P.D.: “Long-Term Production Results of Polymer
Casing Leaks,” Oil & Gas J. (Oct. 13, 1997) 44-46. Treatments in Production Wells in Western Kansas,” JPT (April
28. Jurinak, J.J. and Summers, L.E.: “Oilfield Applications of 1993) 356-62.
Colloidal Silica Gel,” SPEPE (Nov. 1991) 406-412. 48. Kintzele, M.J.: “Polymer Treatments Brighten Production
29. Seright, R.S.: “Effect of Rock Permeability on Gel Performance Profiles of Rocky Mountain Wells,” American Oil & Gas
in Fluid-Diversion Applications,” In Situ (1993) 17(4), 363-386. Reporter (Feb. 1998) 108-119.
30. Krishnan, P. et al.: “Dehydration and Permeability of Gels Used 49. Sydansk, R.D. and Moore, P.E.: “Gel Conformance Treatments
in In-Situ Permeability Modification Treatments,” paper SPE Increase Oil Production in Wyoming,” Oil & Gas J. (Jan. 20,
59347 presented at the 2000 SPE/DOE Improved Oil Recovery 1992) 40-45.
Symposium, Tulsa, April 3-5. 50. Borling, D.C.: “Injection Conformance Control Case Histories
31. Sydansk, R.D.: “Acrylamide-Polymer/Chromium(III)- Using Gels at the Wertz Field CO2 Tertiary Flood in Wyoming,
Carboxylate Gels for Near Wellbore Matrix Treatments,” SPE USA,” paper SPE 27825 presented at the 1994 SPE/DOE
Advanced Technology Series (April 1993) 1 (1), 146-152. Symposium on Improved Oil Recovery, Tulsa, April 17-20.
32. Whitney, D.D., Montgomery, D.W., and Hutchins, R.D.: “Water 51. Lane, R.H. and Sanders, G.S.: “Water Shutoff Through Fullbore
Shutoff in the North Sea: Testing a New Polymer Gel System in Placement of Polymer Gel in Faulted and in Hydraulically
the Heather Field, UKCS Block 2.5,” SPEPF (May 1996) 108- Fractured Producers of the Prudhoe Bay Field,” paper SPE
112. 29475 presented at the 1995 SPE Production Operations
33. Perez, D. et al.: “Applications of a Polymer Gel for Symposium, Oklahoma City, April 2-4.
Establishment of Zonal Isolations and Water Shut-Off in 52. Seright, R.S.: “Gel Placement in Fractured Systems,” SPEPF
Carbonate Formation,” paper SPE 37622 presented at the 1997 (Nov. 1995) 241-248.
SPE/IADC Drilling Conference, Amsterdam, March 4-6. 53. Lane, R.H., and Seright, R.S.: “Gel Water Shutoff in Fractured
34. Seright, R.S.: “Gel Propagation Through Fractures,” paper SPE or Faulted Horizontal Wells,” paper CIM/SPE 65527 presented
59316 presented at the 2000 SPE/DOE Improved Oil Recovery at the 2000 SPE/Petroleum Society of CIM International
Symposium, Tulsa, April 3-5. Conference on Horizontal Well Technology held in Calgary,
35. Lai, Q.J. et al.: “Gel-Cement Combination Squeezes for Gas Nov. 6-8.
Shutoff,” paper SPE 54596 presented at the 1999 SPE Western 54. Tester, J.W., Bivins, R.L., and Potter, R.M.: “Interwell Tracer
Regional Meeting, Anchorage, May 26-28. Analyses of a Hydraulically Fractured Granitic Geothermal
36. Sanders, G.S., Chambers, M.J., and Lane, R.H.: “Successful Gas Reservoir,” SPEJ (Aug. 1982) 537-554.
Shutoff With Polymer Gel Using Temperature Modeling and 55. Seright, R.S. and Lee, R.L.: “Gel Treatments for Reducing
Selective Placement in the Prudhoe Bay Field,” paper SPE Channeling Through Naturally Fractured Reservoirs,” SPEPF
28502 presented at the 1994 SPE Annual Technical Conference (Nov. 1999) 269-276.
and Exhibition, New Orleans, Sept. 25-28. 56. Merrill, L. S.: “Fiber Reinforced Gel for Use in Subterranean
37. Plahn, S.V. et al.: “A Novel Dual Injection System for Water Process Treatment Process,” U.S. Patent No. 5,377,760 (1995).
Shut-Off Treatments,” SPEPF (Nov. 1998) 243-249.
14 R.S. SERIGHT, R.H. LANE, AND R.D. SYDANSK SPE 70067

57. Tank, W. J.: “Short Radius Horizontal Drainhole Development 66. Root, P.J. and Skiba. F.F.: “Crossflow Effects During an
and Application in the Yates Field Unit, Pecos County, Texas,” Idealized Displacement Process in a Stratified Reservoir,” SPEJ
MS Thesis, The University of Texas at Austin, Austin, TX (1965) 229-237.
(1994). 67. Fletcher, A.J.P. et al.: “Deep Diverting Gels for Very Cost-
58. Sydansk, R. D.: “Wellbore Cementing Process Using a Polymer Effective Waterflood Control,” J. Polym. Sci. & Eng. (April
Gel,” U.S. Patent No. 4,724,906 (1988). 1992) 7(1-2) 33-43.
59. Seright, R.S: “Mechanism for Gel Propagation Through 68. Kvanvik, B.A. et al.: “An Evaluation of Stable Gel Systems for
Fractures,” paper SPE 55628 presented at the 1999 SPE Rocky Deep Injector Treatments and High-Temperature Producer
Mountain Regional Meeting, Gillette, May 15-18. Treatments,” presented at the 8th European IOR Symposium,
60. O’Brien, W.J., Stratton, J.J., and Lane, R.H.: “Mechanistic Vienna, Austria, May 15-17, 1995
Reservoir Modeling Improves Fissure Treatment Gel Design in 69. Maitin, B.K.: “Performance Analysis of Several Polyacrylamide
Horizontal Injectors, Idd El Shargi North Dome Field, Qatar”, Floods in North German Oil Fields, paper SPE 21118 presented
paper SPE 56743 presented at the 1999 SPE Annual Technical at the 1992 SPE/DOE Symposium on Improved Oil Recovery,
Conference and Exhibition, Houston, Oct. 5-8. Tulsa, April 22-24.
61. Seright, R.S: “Polymer Gel Dehydration During Extrusion 70. Koning, E.J.L., Mentzer, E., and Heemskerk, J.: “Evaluation of
Through Fractures,” SPEPF (May 1999) 110-116. a Pilot Polymer Flood in the Marmul Field, Oman,” paper SPE
62. Tweidt, L.I. et al.: “Improving Sweep Efficiency in the Norman 18092 presented at the 1988 SPE Annual Technical Conference
Wells Naturally Fractured Reservoir Through the Use of and Exhibition, Houston, Oct. 2-5.
Polymer Gels: A Field Case History”, paper SPE 38901 71. Wang, Demin et al.: “Commercial Test of Polymer Flooding in
presented at the 1997 SPE Annual Technical Conference and Daqing Oil Field,” paper SPE 29902 presented at the 1995 SPE
Exhibition, San Antonio, Oct. 4-6. International Meeting on Petroleum Engineering, Beijing, Nov.
63. Southwell, G.P. and Posey, S.M.: “Applications and Results of 14-17.
Acrylamide-Polymer/Chromium(III) Carboxylate Gels,” paper
SPE 27779 presented at the 1994 SPE/DOE Symposium on SI Metric Conversion Factors
Improved Oil Recovery, Tulsa, April 17-20. bbl x 1.589 873 E-01 = m3
64. Seright, R.S.: “Using Chemicals to Optimize Conformance
cp x 1.0* E-03 = Pa⋅s
Control in Fractured Reservoirs,” Annual Technical Progress
Report (U.S. DOE Report DOE/BC/15110-2), U.S. DOE ft x 3.048* E-01 = m
Contract DE-AC26-98BC15110, (Sept. 1999) 67-75. in. x 2.54* E+00 = cm
65. Faber. M.J. et al.: “Water Shut-Off Field Experience with a mD x 9.869 233 E-04 = µm2
Relative Permeability Modification System in the Marmul Field psi x 6.894 757 E+00 = kPa
(Oman),” paper SPE 39633 presented at the 1998 SPE/DOE *Conversion is exact.
Improved Oil Recovery Symposium, Tulsa, April 19-22.

Table 1—Excess Water Production Problems and Treatment Categories

(Categories are listed in increasing order of treatment difficulty)

Category A: “Conventional” Treatments Normally Are an Effective Choice

1. Casing leaks without flow restrictions.
2. Flow behind pipe without flow restrictions.
3. Unfractured wells (injectors or producers) with effective barriers to crossflow.

Category B: Treatments with Gelants Normally Are an Effective Choice

4. Casing leaks with flow restrictions.
5. Flow behind pipe with flow restrictions.
6. “Two-dimensional coning” through a hydraulic fracture from an aquifer.
7. Natural fracture system leading to an aquifer.

Category C: Treatments with Preformed Gels Are an Effective Choice

8. Faults or fractures crossing a deviated or horizontal well.
9. Single fracture causing channeling between wells.
10. Natural fracture system allowing channeling between wells.

Category D: Difficult Problems Where Gel Treatments Should Not Be Used

11. Three-dimensional coning.
12. Cusping.
13. Channeling through strata (no fractures), with crossflow.
SPE 70067 A STRATEGY FOR ATTACKING EXCESS WATER PRODUCTION 15

Table 2—Water Shutoff Materials and Methods

Chemical & Physical Plugging Agents Mechanical & Well Techniques
cement, sand, calcium carbonate packers, bridge plugs, patches
gels, resins well abandonment, infill drilling
foams, emulsions, particulates, precipitates, microorganisms pattern flow control
polymer/mobility-control floods horizontal wells
Fraction of original productivity

1
Radial flow (unfractured well),
0.8 40-acre 5-spot pattern,
rw = 0.33 ft
or injectivity

0.6
rgel = 5 ft
0.4
rgel = 50 ft
0.2

0
1 10 100
Residual resistance factor
Fig. 1—Fraction of original injectivity or productivity retained versus residual resistance factor.

10000
Volume injected per 100 ft of

0.1 BPM 1 BPM

fracture height, bbls

1000

100
10 BPM

10

wf = 0.04 in.
1
10 100 1000

Distance of penetration, ft
Fig. 2—Gel propagation predictions in long two-wing fractures. Fracture width = 0.04 in.
16 R.S. SERIGHT, R.H. LANE, AND R.D. SYDANSK SPE 70067

10000

Volume injected per 100 ft of

fracture height, bbls
1000

1 i n.
100 wf =
.
0.1 in
=
wf n.
10
.01i
= 0
wf 1 BPM
1
10 100 1000

Distance of penetration, ft
Fig. 3—Gel propagation predictions in long two-wing fractures. Injection rate = 1 BPM.