Formation Damage and Remedial: Stimulation

Download as pdf or txt
Download as pdf or txt
You are on page 1of 17

Clay Minerals (1986) 21, 735-751

FORMATION DAMAGE AND REMEDIAL


STIMULATION

T. R. HARPER AND D. C. BULLER

BP Research Centre, Chertsey Road, Sunbury-on-Thames, Middlesex TW16 7LN

(Received 2 December 1985; revised 18 January 1986)

A B S T R A C T : It is argued that a wide range of engineering data and geologicaldata is required


to make confident prognoses or diagnoses of formation damage. Potential sources of data are
reviewed, with comments on their limitations. It is proposed that microscopicfailure can cause
formation damage in macroscopicallyintact rock at stress concentrations around the wellbore.
The significance of natural fractures is reviewed in relation to formation damage induced by
diagenetic mineralization or invasion by lost circulation material. Removal of such plugging
material by acid stimulationis assessedin relation to diverting agent requirements.

In general, the poorer the quality of a reservoir, the greater the susceptibility to formation
damage. The decrease in quality may relate to a greater content and diversity of diagenetic
minerals, or to lithological complexity and an associated anisotropy and heterogeneity. A
complex diagenetic history may also reflect a relatively complex tectonic history, leading to
the development of small-scale structures (another form of heterogeneity), or to mechanical
characteristics which promote other forms of formation damage.
Any formation damage potential is only realized during drilling, completion and
production operations. Consequently, studies of susceptibility to damage have limited value
if conducted without consideration of the engineering activities during which a damage
potential may or may not be realized.
The practical significance of damage is dependent on the ease with which this can be
avoided or rectified: changes of drilling, completion or production practice can obviate
many forms of damage, whilst stimulation treatments can be used as a remedial measure.
None of these measures can be implemented in a rational and optimized manner without
first identifying the processes which give rise to the damage. Therefore, the practical
significance of formation damage is also a function of the difficulty or otherwise of
determining the cause.
Formation damage results from engineering activities involving geological materials, and
concerns solely the efficiency of fluid flow between reservoir and wellbore. Consequently,
any attempt to predict well performance on the basis of the clastic and diagenetic mineral
assemblage in a reservoir pre-supposes a knowledge of the response of that mineral
assemblage to both chemical changes and to fluid-pressure distributions of the type
imposed by a well. Despite early attempts to predict the characteristics of even single
phase, laminar fluid flow in homogeneous, monomineralic sandstones on the basis of
grain-size distribution (Hazen, 1911), these predictions remain only approximate (Taylor,
1948). The need to conduct fluid flow experiments to determine permeability remains

1986 The Mineralogical Society


736 T. R. Harper and D. C. Buller
universally accepted. This suggests that prediction of the multiphase flow characteristics of
argillaceous sandstones, a far more complex situation, should only be made with extreme
caution in the absence of test data. This is a common engineering situation. The engineering
performanc e of only the simplest system can be characterized by visual inspection alone.
Multiphase flow through porous media cannot be regarded as a simple system.
Consequently, it is of limited benefit to focus on one aspect of the problem in isolation,
whether this be the petrology of the hydrocarbon host rock or the engineering
characteristics of the well which is used to exploit the resource. Predictions based on such
isolated aspects may, of course, be required to meet operational needs. As noted, such
interpretation is complex and the results merit a commensurate degree of caution.
The subsequent discussion of formation damage diagnosis is presented from a broad
perspective of integrated engineering and geological studies. In addition, an aspect of the
more specialized topic of natural fracture stimulation is addressed, the significance of any
natural fractures becoming increasingly important as reservoir quality is reduced, either by
diagenetic mineralization or other means.

CAUSES OF F O R M A T I O N DAMAGE
For purposes of this discussion, formation damage is defined as any process which gives
rise to a positive skin factor. The term skin factor is that introduced by van Everdingen
(1953) and refers to near-wellbore pressure losses over and above that which would arise
during radial flow to a wellbore in an homogeneous, isotropic medium. This is not a
rigorous definition. For example, gas or liquid blocking which would give rise to a positive
skin factor might not be regarded as formation damage; nor would some forms of partial
penetration. Partial penetration could arise because of formation damage, however, where
wellbore enlargement leads to poor communication in perforated completions. Any
definition of formation damage is likely to be controversial, but to pursue this debate here
would be of limited practical significance. We prefer simply to consider those processes
which are relevant to well performance, and which are associated with positive skin factors
observed in well tests. This approach is adopted because of the frequent need to diagnose
the cause of impaired well productivity, and therefore the need to be aware of all possible
causes. Remedial or avoidance measures can only be devised on a rational basis by first
identifying the cause, or potential cause, of damage. In essence, the discussion concerns
factors affecting well productivity impairment rather than the more restricted topic of
formation damage because this approach has greater practical application.
In the most general sense, well productivity impairment can result from five
mechanisms: (i) fluid/rock incompatibility; (ii) fluid/fluid incompatibility; (iii) departure
from laminar, radial flow in an homogeneous, isotropic medium; (iv) mechanical
deformation around the borehole or perforation tunnels; (v) reduction of fluid pressure
during production.

Near-wellbore permeability reduction caused byfluid/rock incompatibility


The influential role of clay minerals on fluid flow in geological materials is widely
recognized. Numerous authors (e.g. Gruesbeck & Collins, 1982) have noted that clay
particle migration and deposition may occur. The circumstances in which this occurs are a
Formation damage and remedial stimulation 737
function of a number of variables. They include the chemistry of the clay mineral(s) and the
chemical and electrochemical nature of both the natural pore-fluid and of the fluid which is
introduced by hydrocarbon extraction operations. Clay swelling may also be induced by
changes of pore-fluid. Other authors have noted that migration of particulate material other
than clays may occur. Muecke (1979) analysed unconsolidated sandstones and observed
that most of the fines present in the pore space were not clay minerals.
It would be imprudent to assume that fluid/rock incompatibility occurs only because of
swelling and particle migration/deposition. Fig. 1 shows the result of a laboratory flow test
in which fluid/rock incompatibility is inferred. A conventional oil-based mud filtrate was
introduced into a reservoir sandstone containing a suite of clay minerals including
mixed-layer illite-smectite, illite, chlorite and corrensite. A substantial loss of permeability
was recorded. Fig. 2 reveals that a reformulated mud filtrate was not damaging to a sample
of the same material. No evidence could be found to ascribe the apparent damage observed
in the first test to fines migration, or to other recognized processes.
Whilst electrochemical forces should certainly not be ignored, an additional factor, the
seepage forces, will also influence particle migration. Seepage force is the mechanical force
acting to displace a particle in the presence of a fluid gradient (which can be extremely high
in the vicinity of a wellbore). The seepage force, expressed as a force per unit volume of the
porous medium through which flow occurs, is given for single-phase flow by iTi, where i is

PERMEABILITY (mD) FLOW RATE (ml/min)

10

I FORWARD
|

6 84 !

5
9-w i !

- I
4-

3-
REVERSE

2-

1-

'
1000 I I '
2000 3000
VOLUME (roll 1
= R E F E R E N C E OIL -!- -!- REFERENCE OIL ---~
$OMF

SOMF = 81MULATED OIL-BASED MUD FILTRATE

FIG. ]. Damage incurred in laboratory core test by flow of off-based mud filtrate.
738 T. R. Harper and D. C. Buller
PERMEABILITY (mD) FLOW RATE (ml/min)

6.0 0.3
5.5 0.25
5.0 0.2
4.5 0.15 FORWARD

4.0 0.1

3.5 0.5

3.0 ml 00

2.5 0.5

2.0 0.1

1.5 0.15 REVERSE

1.0 0.2

0.5 0.25

0.0 0.3
i i i i i i I I i i i i i ! i i
500 1000 1S00 2000
FLOOD VOLUME (ml)
REFERENCE OIL _L_I_ REFERENCE OIL --~
=F-I
SOMF
'#4
SOMF SIMULATED OIL-BASED MUD FILTRATE

FIG. 2. Flow of oil-based mud filtrate through laboratory core: mud reformulatedto avoid
damage.

the fluid gradient and 7yis the unit weight of the fluid. The seepage force therefore depends
on the nature of the fluid flowing in the porous medium (oil, water, gas). A threshold value
representing the onset of particle displacement has been observed by several researchers.
For example, Clementz (1982) has recorded a non-linear result from a laboratory flow test
conducted at a series of pressure gradients (reported in terms of flow rate), which was
associated with the displacement of fines from the core. Apart from the effect of seepage
forces, however, several other mechanisms are known to cause a departure from Darcy
linear relationships.
Damage may also be incurred by the use of drilling and completion fluids which
introduce particulate matter into the near-wellbore region of the reservoir. Also, stimulation
fluids can react adversely and precipitation fronts may be detrimental if the design of a
treatment is inadequate.

Near-wellbore permeability reduction caused byfluid/fluid incompatibility


Not infrequently, serious well productivity problems can be attributed to incompatibifity
of fluids. Incompatibility may arise between introduced fluids and reservoir pore-fluid (e.g.
emulsion blocks produced by mixing mud acid and some crude oils, scale) or between fluids
Formation damage and remedial stimulation 739

introduced into the well (e.g. between various additives in a chemical stimulation package).
Incompatibility of introduced fluids can be avoided by careful treatment design and quality
control. Incompatibility of reservoir fluid and introduced fluid may be controlled by use of
preflush/afterflush techniques and limitation of residence time in the reservoir (but may be
rendered only partially effective by irregular distribution of fluids or reactive rock in the
near-wellbore region).
Introduced solids may also occasionally be chemically incompatible with reservoir
fluids. Fig. 3 illustrates an example of this. In this laboratory test, a granular solid used to
control lost circulation in a series of development wells in the Middle East was mixed with a
simulated formation water from the reservoir in question. Although the degree of mixing in
laboratory and field will be different, the resulting colloidal sludge formed at reservoir
temperature clearly demonstrates an incompatibility between the drilling material and
reservoir fluid.

~i~!i~ii!~'i~i!i~~ii!i!iil
i~ii~~IIi~i~~I!~!ii~ii~i~!ii~'~i~~//~
i!!~!/~ ~i!~' !~i~/~i~i~~i!~i~l~iil~i/~i~i~
ili~ili~!!i!!il i liii~iiii!i~
iiiiiii!ii!iiiiiiiiii+ill i~~ i ii!!
i iiiiiii! , il ii i
~ii~i~i/~~ ~ii~84 iilII~84 !!!ii
~ ~i~/ii~i~iii~i!84 ii, ,, ~!!

FIG. 3. Compatibilityof lost circulation material and formation water (a) at 52~ to simulate
cooling during mud circulation: (b) at 93 ~C (reservoirtemperature).
740 T. R. Harper and D. C. Buller

Departure from radial flow in an homogenous, isotropic medium


Skin factor, as originally derived, is related to the departure from radial flow in an
homogeneous, isotropic medium. A positive skin may arise from a reduction of the area
available to flow and/or a departure from purely radial flow. This could be caused by
anisotropy (typically by bedding so that horizontal and vertical permeabilities differ) or
heterogeneity. Tariq et al. (1985) have presented an analysis of the effect of heterogeneity.
Whilst large scale heterogeneity can cause a skin effect, the scale of heterogeneity which
affects well productivity normally ranges from a dimension which is approximately of the
order of a wellbore diameter, and ranges up to one or two orders of magnitude greater
(Harper & Moftah, 1985). Examples are open and filled natural fractures, continuous or
discontinuous shale stringers, local cementation or inclusions such as clasts or cobbles.
Such heterogeneities should be evident in core, provided they are penetrated.

Mechanical deformation around a borehole or perforation tunnel


Erosion of the wellbore during drilling leads to an excessively thick cement sheath in the
resulting out-of-gauge hole. This limits or precludes penetration of perforation tunnels into
the reservoir, and increases the volume of cement filtrate (which may be incompatible with
the reservoir rock). The development of out-of-gauge holes is a function of rock strength,
drill string behaviour, mud characteristics, drilling time and in situ stress state.
Positive skins may also be generated by the creation of a crush zone around a
perforation. When a jet produced by a shaped charge perforator enters the reservoir rock,
the rock is displaced to one side of the jet. Crushing occurs in some rocks, forming a zone
of low permeability around the perforation tunnel.
Both wellbore enlargement and perforation crush zone formation are purely or primarily
a mechanical form of formation damage. They are well known. There is, however, an
additional mechanical factor which does not appear to have been recognized previously as
a potential cause of formation damage. Around any opening, such as a weUbore or
perforation tunnel, the deviatoric stresses in a reservoir are concentrated so that local
increases of stress difference occur. This may result in failure, giving rise to wellbore
breakouts or sand production, in which case there exist some volumes of the material
which have not failed but where conditions are close to failure. Microcracking and high
levels of distortional strain must be expected in the vicinity of grain contacts in such zones
of stress concentration. It is expected that microcracking may result in a permeability
enhancement, but concentrations of distortional strain (and also, possibly to a lesser extent,
hydrostatic strain) might promote fines release and permeability reduction.
Some preliminary experiments have been conducted to assess whether this expectation
can be supported by laboratory data and whether a more rigorous evaluation is warranted.
Reservoir core samples were flow tested with brine at room temperature and incrementally
removed from the permeameter and subjected to a stress difference for a convenient period
(twelve hours or less) before reloading and continuation of the flow testing. The results
were inconclusive, samples showing both increase, decrease, and no change in
permeability. Fig. 4 shows an example where permeability decrease appears to show some
correlation with exposure of the core to a stress difference. Initially, the core was flow
tested in unconfined conditions. Stresses of I000 p.s.i. (radial) and 4000 p.s.i. (axial and
bedding-parallel) were then applied, followed by another period of flow testing, a period of
Formation damage and remedial stimulation 741

TEST FLUID: 5% N a C I

EFFECTIVE S T R E S S
DIFFERENCE
1. 3 0 0 0 psi (20.7 MPa)
175 -
2. 7000 psi ( 4 8 . 3 MPa)

170-

A
a
++
E
v
>-
k-
2
.d 4'
<C
LU
n-
UJ
0.

150-

145.

9 "" I I I I I
so 100 150 200 2so
FLOW VOLUME. ml x 103

FIG. 4. Effect of stress difference on the permeability of a North Sea reservoir core.

loading at 1000 and 8000 p.s.i., more flow testing, and finally 1000 p.s.i, and 11 700 p.s.i.
at which the sample failed. Overall, the results from the pilot test programme were
sufficiently supportive of the expected influence of deviatoric stress that a more rigorous
series of tests has been planned.

Near-wellbore permeability reduction associated with production operations


A reduction of pressure (and also temperature) is associated with flow to a well. In the
reservoir, this occurs primarily within the near-wellbore region. Associated with this are
well-known adverse effects which can cause well productivity impairment, including the
formation of gas blocks or liquid blocks, waxing, scale deposition, deposition of
742 T. R. Harper and D. C. Buller

asphaltenes and fines migration. The pressure drop may not be entirely detrimental,
however, as the Mohr stress circle is displaced away from the rock failure envelope as the
pore pressure around openings is reduced during production.

SOURCES OF INFORMATION AVAILABLE FOR DAMAGE


D I A G N O S I S/PROGNOSIS
In view of the preceding discussion, it is evident that many different aspects of geology and
engineering influence well productivity. The petrology, including the diagenetic mineral
assemblage, is one aspect, but there are many others which we believe must equally be
considered if one is to optimize the chances of diagnosing or prognosing well problems. The
sources of information upon which such diagnosis can be based are listed below with brief
comments on each. Normally, only some of these data are available, or records are
ambiguous. The ideal situation, in terms of the extent of information available to the
formation damage analyst, rarely if ever pertains. Consequently, a considerable degree of
interpretation and judgement is usually required of the analyst.

Drilling, completion and workover records


These represent the basic record of engineering operations. They form a basis for the
initial identification of possible problems (e.g. drilling difficulties, use of loss agents, nature
of perforation, dirty kill fluid). They also provide the framework for designing laboratory
tests to assess potential damage (e.g. fluid/rock compatibility, fluid/fluid compatibility).

Well tests
Pressure transient analysis is the conventional oil industry method for identifying any
impairment of well productivity, which is conventionally quantified in terms of a skin
factor. As such, well tests are the cornerstone of the information available not only to
detect formation damage, but to quantify the effect. In addition, the productivity index of a
well (or, alternatively, injectivity index), measured during a period of flow, is a clear
measure of well performance, and incorporates the skin factor.

Open hole or production logs


Wireline logs in the open hole can be used to indicate such features as out-of-gauge
wellbores and wellbore breakout, fluid invasion depth, stratigraphy, and natural fractures.
Commonly available, they can be used to check for potential problems such as an oversize
cement sheath. Cased hole logs can be used to some extent to detect flow distribution after
a stimulation treatment, and other factors relating to well performance such as the cement
bond (poor cement jobs may give rise to positive skins because of crossflow behind the
casing).

Production records
Trends of production performance may give clues to progressive changes associated
with damaging processes such as waxing or scaling, or to the effects of workovers.
Formation damage and remed&l stimulation 743

Reservoir petrology
Petrological studies provide a basic description of the porous rock morphology and
mineral content. These must be related to the engineering performance of the rock if the
description is to adopt a more positive role than that of pertinent background information.
While the factors governing the relationship between the type, content, morphology and
distribution of minerals in an argillaceous material and response of that porous medium to
fluid flow and changes of pore-fluid electrochemistry have not been defined, the potential
value of a petrological description is necessarily limited. Many authors have attempted to
classify formation damage potential on the basis of petrological descriptions (e.g., most
recently, Basan, 1985). We regard such proposals as speculative until supported by the
results of laboratory flow investigations.

Detailed reservoir structure


Whilst open fractures are known to assist well productivity, and are associated with
negative skin factors, it is less well known that filled fractures can give rise to a positive skin
(Harper & Moftah, 1985). This is sometimes termed a 'pseudoskin', which may be
mistaken for formation damage induced by another process.

Sedimentary features of the reservoir


A positive skin may result simply from the layering associated with sedimentation.
Either unequal bedding-parallel and bedding-normal permeabilities, or the influence of low
permeability layers on flow to the perforations, may be responsible. Heterogeneities of the
sedimentary sequence other than the latter can also give rise to positive skin factors.

Mechanical tests of core samples


Rock mechanical tests can sometimes be of value, such as to assess rock strength when
wellbore enlargement is suspected, or strength reduction after acidizing a core. A variety of
mechanical property indexing methods are available. Use of a hardness value has the merit
of requiring only small test samples. Table 1 shows a sample of such results obtained after

TABLE 1. An example of the effect of m u d acid on the m e c h a n i c a l characteristics


of an argillaceous sandstone.

M e a n distance from A p p r o x i m a t e strength


core inlet face Hardness reduction
Treatment cm (p.s.i.) %

Untreated -- 5000 --
Untreated -- 4200 --
Acidized 1 1200 76
Acidized 3 3800 23
Acidized 5 3500 29
Acidized 7 5300 0
Acidized 9 5200 0
744 T. R. Harper and D. C. Buller
treating a core plug with mud acid (12% HC1, 3% HF). This core included calcite and
dolomite cements, illite, chlorite, serpentinite and smectite-rich clay.

Character&t&s of reservoir fluids


Incompatibility between reservoir fluids and fluids or solids introduced by engineering
operations can adversely affect flow, such as the sample shown in Fig. 3. If the chemistry
of fluids is known, predictions of probable behaviour can be made. Knowledge of the
reservoir fluids is also necessary to assess the potential for scale, wax, asphaltene formation
or the formation of another phase (e.g. gas blocks in crude oils in reservoirs close to the
bubble point).

Fluid-fluid interaction tests


Predictions of incompatibility can be checked by laboratory mixing tests. However,
relating the results to field behaviour may not be straightforward.

Routine and special core analysis


Differences between core tests conducted at standard temperature and pressure and
those conducted at reservoir conditions are often observed. For this and other reasons,
routine core analysis is of limited application in diagnostic investigations of formation
damage. Reservoir-condition tests, if conducted with sufficient care, are a critical source of
information. However, because preserved core is not identical to virgin reservoir rock,
interpretation may not be straightforward. Moreover, extrapolation from a limited number
of core tests to well behaviour demands caution in heterogeneous reservoirs.
Other laboratory results, such as wettability or pore size distribution, normally provide
valuable input to damage studies. Aside from tests, basic core description, which is
orientated towards mechanisms which may cause damage, is essential.

Fluid-rock interaction
Because knowledge of the relationships between the visual and chemical petrology and
the fluid flow characteristics of the medium is currently incomplete and ill-defined, the
empirical approach of flow testing is critical to formation damage investigations. As with
other techniques, this method is limited, in this case by availability of core and by
disturbance to the core prior to testing. Moreover, the duration of a test programme is
usually not insubstantial in relation to the time scales of exploration and production
activities. As noted in the preceding paragraph, it is appropriate to perform tests at
reservoir conditions.

A CASE HISTORY OF FORMATION DAMAGE DIAGNOSIS


Harper & Moftah (1985) reported a formation damage investigation in the Nubian
Sandstone of the Ras Budran reservoir. All of the sources of information noted in the
preceding section were employed in the diagnosis of formation damage.
Formation damage and remedial stimulation 745

The purpose of discussing this investigation in the present context is to illustrate some of
the misleading conclusions which might have been drawn had not a sufficiently diverse
bank of data been used, and the reader is referred to the cited publication for details of the
overall programme.
Skin damage was noted in well tests to be restricted t o the lower reservoir horizons at
Ras Budran (Fig. 5). The reservoir rock is an approximately 20 md, normally-pressured,
porous quartz-sandstone with pore-tilling kaolinite (Fig. 6). It is medium-grained, with
moderately well-sorted and typically sub-rounded grains of which >98% are quartz. Two
cements are present: quartz, and authigenic kaolinite occurring in typical 'book' form and
comprising ~5% of the rock. Quantities of other clays are insignificant. Very thin, filled
fractures are evident in core.

UNIT 111" 9
UNIT 1I 9
UNIT I 121

11,000 --

3500
11,500 J
DEPTH
[]
FT(M) TVSS [] []

12,000 -

3750

12,500
I i I I I I I
0 4 8 12 16 20 24

SKIN F A C T O R
FIG. 5. Variation of skin factor with depth,

Early in the investigation, the possibility that the observed skin factors were an artifact
of the well test analysis resulting from partial penetration, rather than a damaging process,
was considered. However, additional sedimentological study finally eliminated this
possibility by identifying shale horizons extending across the reservoir, which are assumed
to prevent any vertical communication and consequent partial penetration effect.
Damage incurred by kaolinite migration was suggested at the outset as a candidate
damage mechanism, and considerable effort was expended to test this postulation. The
results of a series of reservoir-condition core flow tests eventually eliminated this
possibility. However, mud acid was found to increase core permeability by a factor of 20 or
more.
On the basis Of the well data then available, a correspondence between apparent skin
damage and the use of oil-based drilling mud, rather than seawater-based mud, was
746 T. R. Harper and D. C. Buller

Fro. 6. Pore-tillingkaolinitein Nubian Sandstone.

observed. Consideration of any changes of reservoir wettability from an assumed


water-wet condition by liberation of previously immobile water around the wellbore and
water-trapping, caused by oil-wetting surfactants in the drilling mud, was prompted by the
observed correlation. Special core analysis tests, both conventional and innovative, were
required before it could be concluded that this phenomenon was unlikely to be responsible
for the observed skin damage. Innovative tests were in part required because the history of
core recovery and accompanying processes were recognized as a possible influence on the
results of conventional special core analysis.
Checks on compatibility of reservoir fluids and drilling/completion materials revealed an
incompatibility of formation water and lost circulation material (LCM) used during drilling
(Fig. 3). However, drilling records revealed that use of the LCM had been localized such
that the inferred damage could not be attributed to this (although use of this agent was
subsequently discontinued).
Formation of strontium sulphate scale where water-based mud was used was predicted
on the basis of the chemistry of seawater and the formation water. Laboratory mixing
correspondingly resulted in the formation of a white precipitate. However, this possibility
was subsequently dismissed on the basis of the lack of correlation of high skin factors and
use of water-based mud. Moreover, drilling and completion records revealed that one well
had been suspended for several years with seawater-based fluid in contact with the
formation, yet subsequent tests revealed negligibly small positive skin factors in this well.
Hole deviation correlated with apparent damage. However, wireline logs revealed that
the possibility that out-of-gauge wellbores had formed during deviated drilling could be
rejected.
Formation damage and remedial stimulation 747

FIG. 7. Thin-sectionof filled fracture.

The filled fractures (Fig. 7), generally of one or a few grain diameters, had a permeability
very much less than that of the surrounding matrix. Because of the orientation of these
small-scale structures relative to the core plugs tested in the laboratory, their influence was
not detected by routine core analysis. From tests devised specifically to assess the influence
of these features, it was determined that they are effective barriers to fluid flow. However,
fracture continuity and orientation could not be determined from the unorientated,
available core.
Rejection of all conventional mechanisms of formation damage, achieved only by using
wide sources of information, compelled a close examination of the potential influence of a
network of low permeability fractures distributed throughout the reservoir. Could these
structures by some means cause excessive pressure loss in the volume of rock in the
near-wellbore region? From this analysis it was concluded, on a qualitative basis, that an
apparent local (near-wellbore) positive skin effect could be caused by a pervasive network
or low permeability inclusions, such as those seen in core, for the following reasons: (i) the
area available for flow to the wellbore/perforations is decreased; (ii) the path length is
increased in the vicinity of the weUbore.
748 T. R. Harper and D. C. Buller
Alternative completion options were proposed on the basis of these conclusions. Despite
the laboratory core observations, use of mud acid was rejected as useless in the presence of
the low permeability fractures which contained little or no kaolinite. Not only does acid
tend to spend on high-surface-area clays in preference to quartz, but the problems of
diversion encountered in placing the treatment were not considered trivial. Therefore, the
proposed completion philosophy markedly differed from that which might have been
adopted had the major permeability increases seen in laboratory core acidizing been
accepted without additional investigation.

STIMULATION OF WELLS INTERSECTING NATURAL


FRACTURES
With decrease of reservoir quality, the significance of any natural fractures to the
production process increases as the ratio of kJk increases, where k = formation
permeability and ki = equivalent permeability of~the fracture. In many reservoirs, the
number of perforation/fracture intersections is limited. If much of the production is
channelled to the wellbore via these few intersections, then it is critical that these are clear
of any plugging material (Murphy & Pearce, 1980). Such plugging material may comprise
invaded drill solids (Kasperskii & Panov, 1971). Alternatively, patchy development of
diagenetic fracture filling could cause near-wellbore plugging. The options for stimulation
are primarily hydraulic fracturing and chemical treatments, usually acidizing. Hydraulic
fracturing of naturally fractured reservoirs is not well understood, and may require the
undesirable use of fluid-loss additive to control leak-off to the fractures. For reasons also of
economy, chemical stimulation may appear to be a more attractive option.
Most natural fractures are of irregular aperture. Flowrate in a fracture can be
approximated by parallel-plate relationships, i.e. flowrate is proportional to the cube of
aperture. Consequently, the invasion of drilling fluid (slurry) into a fracture of irregular
aperture can be envisaged as a sequential filling of channels of successively decreasing
aperture until complete sealing around the wellbore intersection is achieved. Unfortunately,
the aforementioned power law relationship which assists the sealing process also predicts
that clean-up of the plugging material requires special precautions. Assuming that the acid
dissolution front is irregular (because of variability of fracture aperture, irregular
distribution of plugging material about the perforation or other reasons), breakthrough of
acid to the region of open fracture will not occur uniformly, and a localized channel leading
from perforation tunnel to open fracture will first be formed. The resistance to flow of acid
offered by the open channel will usually be far less than the resistance offered by the porous
plugging material (by close analogy with the subsequent analysis). In this case, progressive
enlargement of the channel may be expected, with a rapidly decreasing flow of acid in the
fracture plugging material and a rapidly increasing flow in the channel approximately
according to the above power law.
The necessity for use of a diverting agent is therefore apparent. Use of ball sealers would
in this case merely leave the fractures ineffectually producing hydrocarbon via a narrow
channel connecting the open fracture system to the perforation tunnel. Diversion within the
fracture is required if the acid is to continue to remove plugging material other than in the
immediate vicinity of the channel (or channels) which are the first to form.
Candidate diverting agents could include gels, foams, or slurries which typically include
Formation damage and remedial stimulation 749
particulate salt, benzoic acid flakes or soluble resin particles. Carathers et al. (1984) have
suggested the use of polymer gel for this purpose.
In order to assess the likely efficacy of linear gels as 'in-fracture' diverting agents, an
idealized model was developed. A fracture is assumed to be plugged by an acid-soluble
porous medium. This porous medium can be envisaged to be either drill-or-kill fluid solids,
or a volume of porous diagenetic fracture filling. It is simply termed fracture plugging
material (FPM). The following sequence of events in a gas well is envisaged:
(1) An initial acid stage results in an approximately uniform removal of FPM around
the perforation tunnel, which is assumed to intersect the fracture orthogonally at the
centre of the plugged region. Breakthrough of acid by a single channel is assumed. Acid
then flows preferentially down the single channel, enlarging the channel with
correspondingly decreasing radial flow velocity in the FPM, the majority of which
remains untreated.
(2) A linear gel stage is injected. Flow is restricted to the channel.
(3) A second acid stage is commenced. At this stage of development, the conceptual
model is as in Fig. 8. Additional assumptions include the use of parallel plate
relationships to describe flow in the open channel, and that the fracture does not open or
close in response to any stress changes. Referring to Fig. 8, the relative flow rate of acid
in the porous medium (FPM) and of the acid-gel interface in the channel may be
compared by considering the point at which the second acid stage has flowed to a radial

OPENFRACTURE D
PLUGGEDFRACTURE [ ]
FlG. 8. Idealizedcond~ionsassumedtope~ainduringacidizingofblock~natural~acturea~er
em~acemem ~agelp|uginchannelopenedbyadd.
750 T. R. Harper and D. C. Buller
distance r 1 from the perforation tunnel. Relationships given by Muskat (1946) are
utilized (oilfield units).
For flow of acid in the FPM:

k Pl - P 2
Valr=r ----
rl//a In rz/r ~
For flow of gel in the channel:

kl ( P ~ - P 2 ) 10a ez ( P l - P 2 )
.... = - - .
/tg (r2-rl) 12#g ( r 2 - r 1)
where e, the fracture aperture, is measured in cm, fla and/~g are the viscosities of acid and
gel respectively, and Va and Vg the respective flow velocities.
Thus:

Va -- 1.2 • 10 -7 k /~g r2 - r1
Vg e2 "l~a rl In (r2/rl)
Typical values can be used to demonstrate that the acid flowrate in the plugging material is
negligibly slow compared to the rate at which the gel plug is 'extruded' from the channel.
For example, i f k = ! mD, e = 0.1 cm, #g = 200 cp and/~ a = 1 cp, then:

r 2 -- r 1
V a / V , = 2-4 x 10 -6
r I In (rz/r 0 "
This idealized model therefore strongly suggests that use of a gel (or foam) as an in-fracture
diverting agent is unlikely to permit effective removal of plugging material around the
perforation tunnel. The value of linear gels as diverting agents in fractured reservoirs may
therefore be restricted to diversion from one path to another in open fracture systems.

CONCLUSIONS
Engineering tests are normally required to assess engineering performance, and the
prognosis or diagnosis of formation damage is no exception. In the absence of such tests,
and without the use of all available data encompassing both geological and engineering
activities, conclusions must be treated with an appropriate degree of caution.
It is proposed that failure on the microscopic scale, occurring in rock which has not
failed on the macroscopic scale, can be a cause of formation damage. The nature,
significance and sense of permeability change (increase or decrease) associated with this
process has yet to be established.
Acid stimulation designs are normally based upon the assumption of porous media flow.
This approach is shown to be inadequate in fractured reservoirs, where acid treatments
may more critically influence well productivity than in porous media. Diversion within a
fracture is necessary, and analysis has indicated that viscous linear gel stages are
inadequate for this purpose if the intention of the treatment is to remove porous filling
material.
Formation damage and remedial stimulation 751

ACKNOWLEDGMENTS
The authors are grateful to J. Hyatt and S. Bishop for the reported laboratory results, and to Mr J. F. Berry for
informative discussion. We are pleased to acknowledge the support and encouragement given by the British
Petroleum Company.

REFERENCES
BASAN P.B. (1985) Formation Damage Index Number: a model for evaluation of fluid sensitivity in shaly
sandstones. Proc. Ann. Tech. Conf. and Exhibition of the Soc. Pet. Eng., SPE 14317.
CARATHERS B.K., MILSON S. & TERRACINA J.M. (1984) Polymer gelled block: a diverting agent for acid
stimulations. Southwestern Petroleum Short Course, Texas Technological University, Lubbock, Texas.
CLEMENTZ D.M. (1982) Alteration of rock properties by adsorption of petroleum heavy ends. Implications for
Enhanced Oil Recovery, 3rd SPE/DOE Syrup. Enhanced OiI Recovery, SPE/DOE 10683.
GRUESBECK C. & COLUNS R.E. (1982) Entrainment and deposition of fine particles in porous media. Soc. Pet.
Eng. J., Dec. 847-856.
HARPER T.R. & MOFXAH I. (1985) Skin effect and completion options in the Ras Budran reservoir. SPE
Middle East Oil Technical Conference and Exhibition, SPE 13708.
HAZEN A. (1911) Discussion of 'Dams on Sand Foundations' by A. C. Koenig. Trans. Am. Soc. C.E. 77, 199.
KASPERSKn B.V. & PANOV B.D. (1971) Investigations on plugging by weighted drilling muds on slotted
models. Burenie 5, 28-31 (in Russian).
MtrECKE T.W. (1982) Formation fines and factors controlling their movement in porous media. J. Pet. Tech.
Feb., 144-150.
MtJRPrtY H. & PEARCE R. (1980) Pressure losses in fracture-dominated reservoirs: the wellbore constriction
effect, Geothermal lnst. 1980 Workshop, Univ. of Auckland, preprint.
MUSKaT M. (1946) The Flow of Homogeneous Fluids through Porous Media. Edwards, Ann Arbor, Michigan.
TAmQ S.M., ICHARA M.J. & AYESTA~N L. (1985) Performance of perforated completions in the presence of
common heterogeneities: anisotropy, laminations or natural fractures. Proc. Ann. Tech. Conf. and
Exhibition of the Soc. Pet. Eng., SPE 14320.
TAYLOR D.W. (1948) Fundamentals of Soil Mechanics, pp 112-122. Wiley, New York.
VAN EVERDINGEN A.F. (1953) The skin effect and its influence on the productive capacity of a well. Trans.
AIME 198, 171-176.

You might also like