1/18/2005

SPE 91987
Optimal Production Management in Unconsolidated Mature Sandstone Environment
S.S.Mufarji, MB Petroleum Services L.L.C.

Copyright 2005, Society of Petroleum Engineers Inc.

th
This paper was prepared for presentation at the 14 SPE Middle East Oil & Gas Show and
Conference held in Bahrain International Exhibition Centre, Bahrain, 1215 March 2005.

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

Proposal
In matured unconsolidated clastic hydrocarbon formations
there are several factors which contribute to production
declines. One of the major factors that significantly impairs
inflow into the welbore is solid fines particles which if not
managed appropriately can jeopardise hydrocarbon
productivity hence prevent maximum/plateau production
maintenance.
This paper, presents a new and unique technique which
incorporates a combination of a CFD and 3D reservoir
simulation approaches to study conditions leading towards
pore blocking, characteristics of the damage. Based on the
analyses results, optimization suggestions are given.
This approach eliminates todays core flood limitations which
keep us away from being capable of reducing or avoiding pore
plugging damage. The limitations include: inability to inject
hydrocarbon gas in a multiphase laboratory test, nonrepresentative nature of core sample size, sample conditions
being
different
from
reservoir
condition,
the questionable applicability of scaling up to field extent and
so forth. Hence based on these limitations, there is a need for
better pore blocking characterization techniques in risk
assessments.
Results suggest that in high permeability sand, the finer the
grain size (10 -100microns) the lesser the pressure decline
rate. On the other hand, when 150 - 200 microns fines were
simulated, the decline in pressure was comparatively more.
The sensitivity of grain size in tighter formation is less
compared to that found in more permeable formations as in the
letter, higher magnitudes of permeability decline were
revealed in which case a shallow invasion mechanism was
likely. Unlike the grain size influence to pressure drop and
permeability decline, the simulated fines concentration
ranging from 20% to 50% by volume did not have significant
impact. As far as water cut sensitivity is concerned, significant

difference in permeability drop was observed between 0% and

10% water cut at fixed fines concentration.
Introduction: Causes of Sub-optimal Production
Majority of studies which address reservoir impairment
mechanisms in mature sandstone environment, conclude that
among the major contributing factors are the fines migration
and particle intrusion in formation pores. The possible
consequences in terms of permeability damage caused by fines
movement, accumulation and/ or trapping in pore spaces, and
their impact on pore size reduction is flow restriction, hence
production decline. Mature reservoirs are more affected due to
effects of long term migration and deposition. Similarly, solid
intrusions from the wellbore into the formation especially
during water injection, drilling operations or workover
operations result in injectivity and/or productivity reduction.
Experimental work, analytical approaches and numerical
modelling as well as production test analyses confirmed the
detrimental effects since immediate production declines
occurred when wells were produced under those
circumstances. The migration of clays such as kaolinite, illite,
chlorite, mixed-layer clays and feldspar has been thoroughly
studied and particle sizes analysed. Example1 is given in
Figure-1.

Figure 1: The 100m scale shows mixture of illite/feldspar. The 500m

shows variations in fines grain sizes

Suspended fines particles adhered to pore walls will reduce

pore diameter, and subsequent pore blocking will happen as
particles continue to accumulate (surface deposition). It is also
reported that, fines can migrate in the non-wetting phase1-2
based on several core analyses which revealed that pore
damage associated with fines migration involving migration of

www.petroman.ir

SPE 91987

Production Optimization Guideline

Advancements in slot manufacturing techniques developed by
key slotted liner suppliers, succeeded in making slot sizes that
match or surpass the required sand control used to be provided
by the use of wire wrap screens. This new slotting
manufacturing technique was recently tried in two horizontal
wells. Below is a comparison of initial production tests of
wells completed with keystoned slotted liners (Wells No. 2
and 3) having slot size appropriately distributed with well
No.1 completed with slotted liner without taking into
consideration allocation of various slot sizes along the
horizontal wellbore section. From Figure-2, clearly better
performance is from the wells 2 and 3 having slotted liners
tallied based on expected flow nature with corresponding slot
size per horizontal section, though we can not at this stage
comment on reservoir quality differences from those wells.
During the liner installation trial in Well No.2, slot size was

used and not slot density to promote uniform production

distribution throughout the horizontal section. Although the
effect is slight compared to redistributing slot density on each
joint, it is recommended to place tighter slots in the heel of the
well and larger slot specifications outwards to the toe of the
well.
Keystoned Slotted Liners Performance
Well No.1
Well No. 2
Well No.3
300
250

Oil Rate (m3/d)

kaolinite mineral can take place in both wetting and nonwetting phases. So in both mud types oil-based mud and
water-based, migration occurs and causes pore damage.
In another laboratory study, smectites clay particles are
reported to contain large amount of sodium ions which cause
the mineral to adsorb water and swell several times the
original volume in the presence of fresh water2. Smectite is a
2:1 type of clay composed of two tetrahedral and one
octahedral layers bound together by weak interactive forces.
Smectites and mixed-layer clays are referred to as swelling
clays2. Several crystal structures sometimes form an irregular
fibrous network in a pore space, reducing the capacity of fluid
flow3.
These clay minerals usually occur as loose pore-filling or
pore-lining aggregates in the formation. Solid intrusion into
the formation from the wellbore especially during water
injection, induces damage if particle size control through
proper filtration is not carried out4. For example in mature
sandstone formation, creation of filter-cake during drilling is
associated with filtrate loss of the mud cake. The mud cake is
important in strengthening the wellbore surface thereby
preventing formation caving-in meanwhile minimises fines
blocking of gravel pack, wire-wrapped screen, and slotted
liner. Deeper particle invasion is always detrimental as in that
case it will prevent hydrocarbon fluid movement from the
reservoir to the wellbore or vice versa. The formation of filter
cake at the wellbore wall needs careful considerations when
analysing the ideal particle sizes to be mixed in the drill fluid
systems. In addition in some cases due to incompatibility of
water circulated in the well during completion operations,
precipitates may form whose size may vary significantly. Also
excessive solid and solid type can affect filtrate as it comes
into contact with the formation fluid. To many investigators,
particle grain sizes as small as 0.05m to 7 m can result in
formation damage depending on pore sizes range and flow
velocity as gravity effects are likely to take place in case of
low speed. In a mixture of a fluid and particles (granular) flow
velocity, the smaller particles (sub-microns) can cause gradual
permeability decline while penetrate deeper. Particles
tend to settle down due to gravity effect
or size exclusion mechanism.

200
150
100
50
0
1

10

11

12

13

14

15

16

17

18

19

Well Test No.

Figure 2: Demonstrating production improvement in proper slot

size distribution along wellbore

A study carried out by James R. Tague5, on production

optimization in areas affected by various pore plugging
damage mechanisms, concluded that there are three main
causes of pore damage. The study was conducted using
comparative analysis, laboratory studies, and downhole video
camera. All causes are covered below.
For effective sand control design and evaluation, a summary
of ways of determining pore plugging mechanism is given as
follows6:
If (DG/DF) >15, then "no interaction", will occur since fine
grains can easily pass
through gravel mean diameter.
If (10<DG/DF<15), then "Pore filling", is possible
If (6.5< DG/DF <10), then "Internal bridging/single pore
blocking", may occur
If (5< DG/DF <6.5), then "Shallow internal bridging", may
happen
If (DG/DF <5), then "External cake/ no invasion" may happen
If (DF> (0.33* DP), then "Instant bridging", should be
expected
If ((0.1* DP) <DF<(0.33* DP), then "Invasion & deposition"
are likely to happen
Where:
DG = Gravel mean diameter in microns
DF = Fines grain size in microns
DP = Pore mean diameter in microns
Furthermore a possible source of formation damage is
particle behaviour due to the presence of hydrogen sulphide.
Formation damage causes are sometimes classified as primary
plugging agents that developed during operations, followed by
secondary agents during production or injection. Hence mature
reservoirs having been subjected to several fluid/solid
interaction stages are even more affected.
Diffusion-convection phenomenon provide clearer insight

www.petroman.ir

20

SPE 91987

towards fines migration analysis as the phenomenon

incorporates both fluid and particle behaviours, which could
be transformed.
Diffusion-convection numerically contains mathematical
expression v which denotes superficial fluid velocity and other
variable parameters7. As v o the variables approach
constant values. In the expression, tracer concentration s can
be replaced by suspended fines concentration s and thus
deliverables are fines transportation and plugging tendencies.
To analyse fines movement only, the s parameter in the
convection equation needs not be included to allow fines to
pass while reducing permeability.
The normal distribution theorem which can be used together
has the following properties: It has two parameters, the mean
and the width, which can be estimated from a sample using
few estimator functions.
In the statistical literature the probability density function of
the normal distribution is often denoted by

N(a, ). The standard normal distribution as

explained by Bock8 has zero mean and unit
variance, i.e.

(0,1) = f(x)= (2) 1/2 exp(x2/2) ----------------------------(1)

In the corresponding distribution, the length of a particle lp is
estimated by the product of number of pore throats with radius
r and pore size as follow:
lp = 4 4 / rf4 . ------------------------------------------------ (2)
According to Dullien9, pore size is given as:
=(32/). --------------------------------------------------- (3)

In addition, the liner slotting technique mentioned above,

provides the ability to vary slot density that can be used to
optimize inflow or outflow distributions. The most significant
conclusion reached was that, inflow resistance depends much
more strongly on slot density than on open area. Inflow
characterization was also used to optimize the slot density
distribution promoting more uniform production throughout
the well.
Wennberg et al10 stated that the permeability of a tube is
proportional to its radius r to the fourth power. Assuming that
our porous medium consists of N tubes also assuming linearity
of particle shape, and that fluid does not stop at the plugged
pores, rate of permeability change with time can then be
expanded as follows:

=-

Us0
r4h

4 4

r(r)

--------------(4)

/={[((m-1))p/][rf/rh]4Us0
((rf4/(44[(32/)])()}-----------------------------------(5)
(m-1))p/ =

q = [7.08103ih(Dp)/((ln( r e /r w)0.75+S))]0.159--------------------(6)
In a research conducted by Shell, result suggests that tests
exceeded the thick-walled cylinder (TWC) stress limit
resulted in initial perforation failure. In other words, initial
production will always cause excessive fines production if the
near-wellbore vertical stress (vwf) exceeds the laboratory
measured stress using a thick-walled cylinder stress (TWC)
system.
Production Optimization through Fines Grains
Management
Using computational fluid dynamics designed porous medium,
simulated fluid pressure decline was found so sharp when
mixed with finer grain sizes between 10-100microns, whereas
as the grain size increased the impact of mixture of coarser
grain sizes, was minimal on pressure. A possible shallow
invasion damage mechanism could be the case.
It is recommended that any slot width tolerance on the low
side of specifications be run in the heel of horizontal wells in
clastic environment. This will help reduce the risk of sand
production in the heel of the well where the greatest draw
down occurs.
Unlike the grain size influence to pressure drop and
permeability reduction, the simulated fines concentration
ranging from 20% to 50% of crude volume in high
permeability sand along horizontal direction, did not have
significant impact. However, when simulations were carried
out across the width of the sand body, clear concentration
effects were observed.
As far as water cut sensitivity is concerned, significant
difference in permeability drop was observed between 0% and
10% water cut at fixed fines concentration.
Among the major outcome of this study work is the ability to
analyse solid fines velocities separately from liquid velocities
while both flow simultaneously.

dr

In this case rh is some length scale put to keep the units simple.
Combining and re-arranging equations according to Al
Mufarji et al7, we arrive at fines migration and clogging
model. Hence the model which computes rate of permeability
change is;

Where:

(32/) =

Given initial permeability , average grid permeability k can

be determined from other equation.
Then from equation 5, rate of change of permeability is
computed.
In field application, the flow equation used is the semi-steady
state phase flow equation:

Computational Model Set-up

More complex quadrilateral and hexahedral grids created by
the pre-processor GeoMesh were used for sensitivities. Once a
grid has been read into the solver, all remaining operations
were performed within the solver. The operations included
setting up of boundary conditions, defining fluid properties,
executing the solution, refining the grid, viewing and postprocessing the results. Fines effects in multiphase flow in a
porous medium modelled with sandstone reservoir properties
were analysed. The porous medium whose dimensions are
60cm (30 cells) by 95cm (50 cells) was further subdivided to

www.petroman.ir

SPE 91987

represent all ranges of properties found in multiphase

reservoirs. The medium was modelled such that it contained
conventional high permeability clastic reservoir properties of
rock and fluids. The fines grain sizes were chosen based on
standard mesh sizes used in the sand control measures.
Various observed pressure profiles in various simulated
reservoir dimensions were recorded. Initially pressure
decreased as fluid entered the porous medium then increased
as it exited the medium into open space since the pressure was
being exerted throughout the period. An observation which
was repeated in all simulations is that the fines velocities were
always slower than crude velocities though both started at the
same speed. The most important fines velocity is the porous
medium average velocity. The study also revealed statistical
results of number of cells penetrated at average velocity. In
this case the average porous medium fines velocity simulated
was 0.066384 m/s which correspond to 1797 cells penetrated.
Decline in fluid pressure was high, from 1.50E8 Pa to 6.50E8
Pa resulting in 5 E8 Pa pressure drop.
Traditionally, simulation engineers assign a constant value of
permeability during the initialisation stage of simulation
studies when dealing with perceived homogeneous reservoirs.
Hence such a simplification is very likely to result into
erroneous reservoir performance predictions. This error
becomes more significant when modelling pressure sources
which do not communicate with each other, referred to as
aquifer cells. As a default, the size of each aquifer cell is a
fraction of the parent aquifer size, the fraction being
proportional to the product of the grid block permeability and
the area of the grid block-aquifer cell interface. This default
may give optimistic results for the aquifer support in cases
where the permeability of the aquifer cells varies strongly
from cell to cell. Variations may be either from the parent
aquifer or from the grid block to which it is connected. As
such the properties of each aquifer cell should be edited
accordingly in reservoir simulators.
Fracture and Matrix Permeability in Mature Sandstones
The reservoir simulators well management
package not only covers network and
interactive editing, but also include
scenarios related to formation compaction
effects
on
reservoir
properties.
For
example the effect of compaction on
permeability can be simulated in both
fractured reservoir scenario and nonfractured setup. The later was used in
this mature sandstone environment study to
avoid
multiple
effects
which
could
otherwise
mask
effects
of
fines
on
reservoir performance. Both fractured and
matrix nodes could be modelled in a
reservoir computational grid.
In a nested, hierarchical way, for each
grid block, the centre of gravity zcoordinate was determined as follows:

Z(r) = rB(r) z(r) dx dy dz + zorigin-------------(7)

rB(r) dx dy dz

nq nq nq

z(r(qi,qj,qk)).det jr (qi,qj,qk) . pipjpk------(8)

i=1 j=1 k=1

v(r)
A block of 40 by 8 by 8 grids for each run was modelled. Also
possible to simulate, is a step change in production say per
year or per any time frame.
Conclusions
In mature sandstone reservoirs, physical fines migration and
solid intrusion damage mechanisms were analysed to
investigate reasons behind retarded fluid mobility from the
reservoir to the well-bore and vice versa.
For fines migration to take place and result in formation
impairment, the fines particles need be dislodged from their
location in the pore surface and entrained in a flowing fluid
through pore structures until they get trapped in pore
constrictions such as pore throats. Various types of damage
mechanisms can occur depending on many factors which
range from the nature of depositional environment of the
reservoir, right to well operations during the life cycle of the
well.
Similarly, solid intrusions from the wellbore into the
formation especially during water injection or drilling
operations can plug pore structures if solid filtration of
injected fluid is not done appropriately. The effects of solid
intrusion and filtrates can be detrimental to the injectivity
performance as they can penetrate right into the formation
affecting its permeability and porosity. As such both
injectivity as well as productivity can be affected.
In addition, fines migration can manifest itself in several ways
including plugging of pore spaces, blocking producing slots,
clogging gravel-packs, and reducing pump efficiency by
eroding plungers, impellers and valves.
Nomenclature
lp = length of particle migration
= statistical parameter
= pore size (m)
rf = Fluid radius
=constant, characteristic of fluid
= grid block average porosity (fraction)
/ = rate of permeability change
Uso =derivative of the tracer concentration
= pore velocity
p =differential pressure across the permeable zone
= fines concentration suspended in the fluid
Dp = pore mean diameter in microns
h = reservoir height
r w = well-bore radius
r e = external radius
= porosity
DG = gravel mean diameter in microns

www.petroman.ir

SPE 91987

DF = fines grain size in microns

= specific surface area (1/L)
vwf = expected near-wellbore vertical effective stress at
perforation depth
TWC = thick-walled cylinder strength.
n = total number of layers perforated
v = superficial fluid velocity
S = Skin factor
References:
1.

Poynton, N., R. Tidswell, Baker Petrolite, and J.

Steele. Squeezing Aqueous Based Scale Inhibitors
into a Water Sensitive Reservoir Development of a
Squeeze Strategy, 2000, SPE Paper No. 60219.

2.

Recham, R. Super-Critical Rate Based on Economic

Recovery in Water and Gas Coning by Using Vertical
and Horizontal Well Performance, 2001, SPE Paper
No. 71820.

3.

Schechter, R.S., Oil Well Stimulation. Englewood

Cliffs, NJ: Prentice Hall. 1992.

4.

Gruesbeck, C., and R. E. Collins : Entrainment and

Deposition of Fine Particles in Porous Media, Soc.
Pet. Eng, J 22. 1982. p.p. 847-856.

5.

Tague J.R. Overcoming Formation Damage in Heavy

Oil Fields: A Comprehensive Approach, 2000, SPE
62546.

6.

Oyeneyin, M. B., J.M. Pedden, Ali Hosseini, and G.

Ren. Factors to Consider in The Effective
Management and Control of Fines in High
Permeability Sands, 1995, SPE Paper No. 30112.

7.

Al Mufarji, S.S., M.B. Oyeneyin, and S. Jihan.

Optimizing Production from Fractured Damaged
Reservoirs, 2004, SPE Paper No. 91971.

8.

Bock R.K..Normal Distribution Theorem, (April

1998).

9.

Dullien F. A. L. Porous media: fluid transport and

pore structure. 2nd ed. San Diego: Academic Press,
1992.

10. Wennberg., E. Kjell, G. Batrouni, A. Hansen.

Modelling Fines Mobilization, Migration and
Clogging, 1995, SPE Paper No. 30111.

www.petroman.ir