SPE

Society of Petroleum Engineers

SPE 21439

Water Quality Requirements for Fractured Injection Wells

P.J. Clifford, * D.W. Mellor, and T.J. Jones, BP Research
SPE Member

Copyright 1991, Society of Petroleum Engineers, Inc.

This paper was prepared for presentation at the SPE Middle East Oil Show held in Bahrain, 16-19 November 1991.

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 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. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment
of where and by whom the paper is presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836 U.S.A., Telex, 730989 SPEDAL.

volume injected per unit area of sand face over the

ABSTRACT lifetime of the well. Information about the pore
structure of the formation rock is also obtained,
Field measurements of water quality, together with and is assumed to correlate with its tendency to
an assessment of the role of thermally induced plug due to particles of a given size 3 ,4. On this
fractures, indicate that stringent fine filtration basis, a filtration requirement is specified.
requirements are often not required for water Depending on the properties of the formation, such
injection wells. Corrosion and scaling of tubulars specifications are often very stringent.
are shown in some cases to result in a bottom-hole
quality worse than raw intake water. VeIl This approach was adopted in specifying filtration
performance, however, is not seen to deteriorate requirements for seawater injection into the
significantly as a result. The reasons for this Forties field (offshore UK) in the mid-1970's.
are examined by the inclusion of fines in models of Similar criteria were also used for the Magnus
injection well fracturing. (offshore UK) and Ula (offshore Norw~y) fields in
the 1980's. Mercury porosimetry was used to
measure a pore size distribution, and the effective
INTRODUCTION hydraulic radius was obtained as a function of
permeability. By choosing cores in the lowest 10%
of the permeability range, and assigning an
The control of water quality is recognised as being
important in preserving the injectivity of wells arbitrary factor of 2 between pore and particle
size to account for bridging, a target upper limit
and the integrity of the injection system 1 for particle size was fixed.
Specifications for the filtration of injected water
have frequently been set based on the need to For Forties field, the design specification was 95%
prevent plugging of the injection sandface by removal of particles greater than 5 ~m in diameter,
suspended fine solids or oil droplets. The with 100% removal of particles larger than 10 ~m,
assumptions underlying the established approach calculated on the basis of a 20 ~m hydraulic
include:
radius. Core injection studies confirmed that
(i) that the bottom-hole water quality is similar face-plugging would occur when larger particles
to the quality immediately downstream of the were injected, and even suggested that the target
filters; limit might be excessively high, in that some
blockage occurred eventually even when all
(ii) that the injected water enters the formation particles above 2~m size were removed from the
through the surface of the wellbore or perforation seawater. The fine filter specification in Ula
tunnels. field was even more stringent than Forties,
requiring 98% removal of particles larger than 2
~m.
Laboratory experiments are carried out to measure
the permeability damage caused to core-plugs by
injection of large volumes of water of given solids Hore recently, however, our experience in water
content 2 The results are extrapolated to the injection has cast doubt upon the need for such
stringent fine filtration specifications in a
number of water injection projects. In particular,
References and illustrations at end of paper it is demonstrated in this paper that surface and
downhole samples show a deterioration in post-

851
2 WATER QUALITY REQUIREMENTS FOR FRACTURED INJECTION WELLS SPE 21439

filtration water quality downstream from the and free chlorine took place continuously pre-,
filter. This deterioration is thought to be due to post- and during downhole sampling. Injection
corrosion and scale products, and in some instances water samples of 650 ml volume were collected from
results in injection of water whose quality is a point just above the perforation tunnels using a
significantly worse than the raw intake quality. downhole sampler, operated via an electric
wireline.
This finding alone is surprising, since it suggests
that wells have been injecting for extended periods Downhole sampling is a technique which requires
at a water quality outwith of specifications, great care in order to obtain a fully
without a significant effect on formation representative water volume. A number of
injectivity. Such an insensitivity to water precautions were taken. The downhole sampling tool
quality is confirmed by the performance of wells was stripped down and cleaned prior to use. The
which have been injecting for periods of time at a sample chambers were then rinsed with filtered
reduced level of filtration. It has even proven water. After the tool had been placed, between 3
possible in some fields, at an adequately high and 12 wellbore volumes of water were injected at
well-head pressure, to re-inject produced water normal injection rates before sampling, in order to
with an oil and solids content grossly in excess of avoid contamination with particles dislodged from
most seawater samples. the tubulars during lowering of the tool. On
removal from the well, the external surface of the
The explanation for this paradox - high tolerance tool was cleaned, the exit port was rigorously
of formations to relatively poor quality water - washed with filtered water and the first 100 ml of
lies with the role played by fractures in the sample was flowed to waste. The remaining sample
injection process. In general, water injection was displaced into a clean glass bottle and sealed.
itself causes significant deformation to the Up to 3 downhole samplers were run on the same tool
formation around the wellbore. This issue has string,and samples were separately collected and
previously been raised in discussing the relevance analysed for comparison.
of laboratory core-flooding to field performances,
and has been recognised in some previous field The sampling tool and method used were considered
considerations of water quality 6. Mechanisms are to be the best available. It is recognised that
discussed in this paper by which most wells in there remain possible sources of contamination in
consolidated sandstone formations which inject downhole samples, for example from contact between
water significantly cooler than the reservoir the wireline and the tubing walls. However. the
temperature will become fractured. Field consistency of results between different wells and
measurements provide widespread confirmation of reservoirs indicates a systematic rather than a
fracturing in formations of this type. random mechanism acting to determine the downhole
quality.
The existence of fractures, sometimes with a A Model TAIl Coulter Counter was used for particle
dimension of the order of 100 m, creates a greatly size analyses. (It should be noted that downhole
increased area of sand face through which the samples must be analysed immediately the sample is
injected water enters the formation. This allows vented to the atmosphere. This is because exposure
the well to tolerate a significantly lower quality of the sample to air causes oxidation of dissolved
of water. However, it is then important to iron, to give a large number of particles in the
investigate what is the response of a fractured size range up to 3 ~m.) A sample volume of 0.5 ml
well to a given loading of suspended solids, both with use of a 100 ~m orifice tube gave particle
in order to set amended water quality limits, and counts in the range 1.5 to 50 ~m. Suspended solids
to see if there may be any further undesirable loadings were also measured, using 0.5 or 0.8 ~m
consequences, due for example to excessive fracture millipore filters, which could be placed in holders
growth. For this purpose, numerical simulations on the topside sample points.
have been carried out, and are described, which
incorporate plugging due to fine solids into our In certain cases, the elemental composition of
model of thermally induced fracture propagation. solids filtered from downhole samples was analysed
qualitatively using the technique of energy
dispersive analysis with X-rays (EDAX) on an
electron microscope.
YATER QUALITY MEASUREMENTS FROM THE FIELD
Results of Analysis
A series of water quality measurements in field
injection systems was carried out with the
objective of determining whether fine filtration Yater quality measurements carried out on wells
was successful in achieving the specified quality injecting seawater into the Ula field demonstrated
significant deterioration in quality downstream of
of water at bottom-hole. As will be shown, the the fine filters. The Ula water injection system
consistent finding of several such studies in North is shown in Figure 1, which indicates the locations
Sea reservoirs was that fine filtration had little of the topside sampling points. Downhole samples
effect on downhole quality. were collected from wells A-03 (flowing at 13700
Analytical Procedures bbl/day) and A-OS (9900 bbl/day). The completions
are constructed from carbon steel. Measures are
taken to reduce oxygen levels in the injected
Yater quality measurement was carried out at water, with use of oxygen scavenger. An oxygen
several points in the injection system: at the sea level of 20 ppb was in fact measured in the
winning pumps, just upstream of the fine filters, injection manifold during the period of water
just downstream of the fine filters, in the quality determination. A dual media fine
injection manifold and at bottom-hole. Topside filtration system was in operation over the entire
monitoring for particle counts, suspended solids period of measurement.

852
SflE 21439 P.J.CLIFFORD, D.W.MELLOR, T.J.JONES 3

Data from gravimetric analysis of total suspended has also been obtained from the seawater injection
solids, shown in Figure 2 and listed in Table 1, system of the Magnus field.
illustrate the trend towards lower quality in
downstream locations. A solids loading of 3.5 mg/l The general result of these measurements is that
upstream of the fine filters is reduced to 0.1 mg/l downhole water quality is significantly worse than
after filtration, which shows that the fine intake water quality, even when fine filtration is
filtration system was operating efficiently. working effectively. Under field conditions of
However, even at the injection manifold, the oxygen control, corrosion/erosion products from the
suspended solids have increased to 0.3 mg/l and ~ubula:s a~d particles in all size ranges,
downhole samples from Ula wells A-05 and A-03 have including Sizes above 10 ~m, to the injected water
loadings of 6.0 and 7.5 mg/l respectively. These during its passage from the well head to the
values are twice as high as the fine filter inlet, perforation tunnels. Removal of fine filters does
and ten times as high as average raw seawater from not appear to have a significant effect on downhole
the North Sea. quali ty.
Detailed particle size analysis, shown in Figure 3
and listed in Table 2, confirms this trend. The
fine filters remove almost all particles of size THERMALLY INDUCED FRACTURING OF INJECTION WELLS
greater than 2 ~m (meeting design specifications),
but downhole samples contain quantities of 16 ~m + Vater injection wells commonly become fractured
particles which are 10 times as great as the number under standard operating conditions, without the
upstream of the filters. These are potentially the need for a deliberate programme of hydraulic
most damaging particles, since they are large fracture stimulation. The condition under which
enough to block pore throats completely. fracturing can occur is that the bottom-hole
pressure in the wellbore should exceed the
Comparable results were obtained from analysis of compressive stress in the formation rock. Before
samples from the seawater injection system of the the reservoir is developed, the gradient of minimum
Charlie platform of Forties field. These are principal horizontal rock stress with depth
significant in that they provide data directly (fracture gradient) will typically lie in the range
relevant to the field injection trials described 0.7-0.8 psi/ft, which will normally exceed the
below in which fine filters were removed. In this injection pressure gradient.
case, downhole samples were taken in wells FC4-1
(injecting at 27000 bbl/day) and FCI-2 (41500 However, there is often a large temperature
bbl/day), both at periods when the fine filters difference between the formation and the injected
were in line, and when they had been removed. water. Vhen water significantly cooler than the
reservoir is injected into it, contraction of the
The solids loading, shown in Figures 4 and 5 as formation will result. This reduces compressive
measured pre- and post-removal of the fine filters, stresses' and allows fractures to propagate at
is 10 times as high downhole as it is downstream of bottom-hole pressures significantly lower than
the filters, for both wells and at both times. The conventionally estimated fracture gradients (by up
fine filtration has an observable but relatively to 0.2 psi/ft in a consolidated sandstone).
small effect on water quality immediately above and
below the filters. The small effect of the filters Depletion of pressure within a reservoir also has
may be attributed in part to the fact that all the effect of reducing the rock stress, and for
measurements were carried out during a period free this reason it is easier for an injection well to
from algal bloom, when the intake water was fracture into a depleted formation than a highly
therefore cleaner than average. Filtration had a pressured formation.
negligible effect on downhole water quality for
both wells. Removal of the fine filters did not Evidence for the existence and behaviour of
lead to an increase in solids loading, and thermally induced fractures has been reported from
additionally did not lead to any reduction in a number of fields 6 S - 11 Fracturing always serves
injectivity during the period of measurement. to increase the injectivity of the well, and for
that reason may be a desirable outcome. However, a
Detailed particle size analysis, shown in Tables 3 large fracture may influence the vertical and
and 4 as measured pre- and post-removal of the fine horizontal sweep performance of the well 12 in ways
filters, shows a significant increase in particle which mayor may not be desirable.
.numbers downhole for all particle size ranges. In
all cases, many particles of size greater than 8 um There will also be substantial thermal and pore-
entered the water downstream of the water injection pressure effects, but of slightly different
header and therefore presumably during its passage character, in other types of formation, including
down the injection well tubing. EDAX analysis of unconsolidated sands and naturally fractured
the downhole samples indicated that iron was the carbonate reservoirs. Poorly consolidated
main component of the suspended solids, with a materials close to an injection wellbore will tend
significant peak in sulphur. This suggests a to suffer failure and disaggregation. The likely
reaction between bisulphite oxygen scavenger and effect of cooling on a naturally fractured system
the pipework, resulting in iron sulphite will be to increase the aperture of existing
precipitation. The downhole particle concentration cracks, rather than creating new fractures. In
may also be associated with an imperfect control of this paper, we deal particularly with the case of
oxygen levels under field conditions, leading to consolidated sandstones.
corrosion of tubulars.
Thermally induced fracturing is of great importance
A very similar result, with downhole water quality to the question of water quality requirements. The
approximately 10 times worse than seawater intake, presence of fractures provides a much larger

853
4 WATER QUALITY REQUIREM~NTS FOR FRACTURED INJECTION WELLS SPE 21439

surface area for leak-off than a non-fractured where C is a dimensionless constant, and L is the
wellbore. Furthermore, the fracture system is able cumulative flux in units of length (m 3 of injected
under some circumstances to grow to expose new volume per m2 of fracture face). The constant C is
sand face and accommodate further downhole determined from core-flood experiments, and depends
filtration. on the water quality and the formation properties,
most notably the permeability.
Note that more complex plugging materials, such as
MODELLING THE EFFECT OF INJECTION FINES ON YELL oily sludges or gels, will not show the simple
PERFORMANCE behaviour described by equation (2). In a previous
publication 13 , the case of plugging by EOR polymers
The Fracture Simulator was studied, with the use of a modified equation
adapted from laboratory core-flood experiments.
The growth of thermally induced fractures from an
injection well is determined by the stresses which Ye do not report any new core-flood results for
are set up by the temperature and fluid pressure seawater injection In this paper. However, the
distributions in the region of the reservoir close significance of the formation damage constant C in
to the well. At the same time, the temperature and equation (2) may become clearer by reference to
pressure distributions are affected by the flow of core injection tests such as those reported by
water from the fracture into the formation. As the McCune 2 A typical test involves injection of
fracture grows, it cools a larger area, which may several thousand pore volumes of contaminated water
promote further fracture growth. The processes of into a core-plug of one inch (2.5 cm) length. If
fluid flow and fracture growth are therefore the effective core-plug permeability is found to be
strongly coupled. reduced by 50% for 500 pore volumes of injection,
then this corresponds to C = 0.01.
The simulation model reflects this coupling by
combining a 3-dimensional finite difference Many of the tests with fine filtration reported by
reservoir simulation of fluid and heat flow in the McCune 2 have C values less than or equal to 0.01.
region around the well with a 3-dimensional Tests with broken- rather than cut-faced cores
boundary element fracture mechanics calculation of reported by Todd et al. s show values of C as low as
fracture growth. At each timestep during the well 0.0001 in some cases. It is only when suspended
history, the pressure, saturation and temperature particles larger than 3 ~m are injected (see Figure
are calculated in the gridblocks of the reservoir 12 of McCune 2 ) that C rises to significantly larger
model, using the fracture as the fluid source term. values of up to about 0.1. It should be noted that
At intervals, the stress state in the plane of the a water with this degree of plugging would
fracture is calculated (using a displacement conventionally be regarded as highly unsuitable for
potential method) and the fracture size is updated injection.
so that it is in equilibrium with the new stress
field. Our in-house studies have indicated that seawater
as injected corresponds to C in the range 0.01 to
A more detailed description of the model, and of 0.1. Yater injection at a solids loading of 10
its application to certain field problems, is mg/l, in a particle size range typical of downhole
provided in earlier publications 12 conditions, into a core of 100 md permeability
will pessimistically yield a value of C of th~
Representation of Injected Solids order of 0.1.

Injection fines are represented by the gradual Effect of Face-Dlu~~in~ on Fractured Yell
build-up of a thin layer of low permeability skin InJectivity and Fracture Growth
at the fracture face, which may be either on the
surface of the fractured rock, or just within the The water injection process has been modelled for
pores. If at some time the skin layer has a wells in a range of formation permeability
thickness dsk and a permeability ksk on some part distributions, with water of different qualities
of the fracture face, the pressure drop across it injected under different well operating
in that region will be: constraints. Results confirm that injectivity
losses due to fines are often small, while showing
(1) that fracture dimensions can sometimes be
significantly affected by water quality.
where (O/A) is the flow rate through unit area of
the fracture face (flux) and ~ is the water Ye consider as our base case a layered sandstone
viscosity. formation at a temperature of 200F (93C) with
seawater injection at a temperature of 60F (16C).
The build-Up of the skin layer on any region of the The permeability structure, identical to that used
fracture face is assumed in this model to depend on for the previous polymer studies 13 , is given in
the cumulative flux of injected water through that Table 5. The vertical permeability is 1% of the
region of the face. For relatively low horizontal permeability. The rock has a minimum
concentrations of fine solids in seawater, the principal in-situ stress of 8000 psi and a
simplest model which describes laboratory core thermoelastic constant, which determines the
thermal stress change, of 15 psi/oF (27 psi/DC).
floods is a skin damage which is directly The poro-elastic constant, which determines the
proportional to the cumulative flux. The face-
stress change due to pore pressure, is taken to be
plugging is then described by the equation: 0.7 psi of stress change per psi of pressure change
(2)
from the initial reservoir pressure of 5000 psi.
The injection well is placed in as-spot

854
SPE 21439 P.J.CLIFFORD, D.W.MELLOR, T.J.JONES 5

configuration with well-spacing of 2100 ft. To investigate a more difficult case for dirty
Producers are operated at a fixed 3000 psi bottom- water injection, a case of a lower conductivity
hole pressure, and the injection well is operated fracture was considered. In this case,
at a fixed rate of 20,000 bbl/day over a 2 year corresponding to turbulent water flow through a
period. fracture of narrow aperture (as would be associated
with high formation stiffness), there is a more
Ve first consider a case without any face-plugging. significant pressure gradient required to force
As cooling advances into the high permeability water through the fracture. A pressure gradient of
layers of the reservoir, the fracture grows into 2 psi per foot of length is used here. This
these layers, according to the calculation pressure gradient reduces pressure at the fracture
illustrated in Figure 6. In this calculation, the tip and significantly restricts horizontal fracture
fracture is approximated as having a high growth in a case without plugging, as shown in
conductivity to internal flow. Its length is Figure 10. In addition, high values of the
restricted initially by the extent of cooling, and formation damage constant C, of up to 1.0, were
later by the fixed rate constraint, though some considered.
growth continues at late times to compensate for
the advance of the low mobility fluid bank of cold For C = 1.0 (an order of magnitude worse than the
water. The bottom-hole pressure, shown in Figure downhole samples), Figure 11 shows that an
7, falls slowly towards about 6300 psi, additional bottom-hole pressure of 350 psi,
corresponding to the stress in the cooled layers. compared to the undamaged case, is required in
Note that the sudden drop in pressure at intervals order to inject at 20,000 bbl/day. (C = 0.1
is a result of the numerical method, and requires an incremental 100 psi.) Again, and in
corresponds to a re-calculation of the fracture spite of the lower fracture conductivity, the rise
size. The values of pressure immediately after a in pressure is alleviated by the growth of the
fracture calculation are the most accurate numbers fracture (Figure 10) to more than double its
to use. original surface area, as a result of face-
plugging. Figure 12 shows that for C = 0.1, there
The effect of face-plugging on this well is small, is a significant diversion of flow, and hence of
even when the formation damage constant C is taken new face-plugging, into the lower permeability
to a value of 0.1, which is typical of the downhole layers.
quality discussed above. The effect on bottom-hole
pressure requirement to maintain injection at the This case demonstrates that a significant
constant rate (Figure 7) is no more than tens of deterioration of well injectivity is not expected
psi after 2 years. If the fracture had not been until the solids concentration reaches several
allowed to grow (looking at for example the times the levels recorded downhole, even in the
intervals in Figure 7 between fracture re- case of fractures which are restricted in length by
calculations) the pressure might have risen by up flow considerations.
to about 200 psi/year at C = 0.1, but in fact what
happens is that in order to maintain the constant Other Effects of Solids Injection
rate, the fracture gradually grows to expose fresh
sandface. This is achieved with only a small Alternative models of suspended solids behaviour
requirement for additional pressure. The effect of may be envisaged in addition to face-plugging of
face-plugging on fracture growth is illustrated in the type which would occur for I-dimensional
Figure 8, and is seen not to be large in this case. injection of water into the sandface. Vater flow
within a fracture is both rapid, with speeds
Figure 9 displays the distribution of skin damage parallel to the fracture face of up to several
(defined as dsk/k sk in equation (2), in units of metres per second, and often turbulent. In such
ft/md) over the fracture face. Ve plot the amount conditions, it is reasonable to regard the problem
of skin damage which builds up between 500 and 700 as being one of dynamic filtration (as is widely
days, for the fracture with C = 0.1. The structure recognised, for example, in the case of drilling
of the diagram corresponds to the grid-block muds) rather than static filtration. As yet,
structure of the reservoir model. It is seen that however, little suitable laboratory evidence exists
the new skin damage during this period is occurring to characterise models of the dynamic filtration of
predominantly close to the fracture boundary, which low concentration suspended solids under the
reflects the fact that more flow is occurring conditions of a fracture interior.
through the newly opened sand face at the boundaries
than in the partially plugged older area in the Additional factors will arise in the case of
fracture centre. injection of highly contaminated produced water, or
for example the disappearance by corrosion of a
A large, high conductivity fracture can maintain significant fraction of the mass of a tubular, as
well injectivity very easily against face-plugging appears to have happened in some cases over a
at water quality typical of measured downhole period of years. These cases could lead to the
seawater. Many water injection wells are believed injection into a fracture of a total volume of
to have fractures in excess of tens of metres in solids comparable to or greater than the volume of
size, and we do not therefore expect their the fracture itself (often not more than a few
injectivity to be affected by solids in seawater, cubic metres). Clearly, these situations impose
with or without fine filtration. It is also not different requirements upon the simulation of
possible for seawater solids to force excessive fracture behaviour.
fracture growth to a length (e.g. 300 m) which
might damage areal sweep performance. This is
because the sand face is then far too large to be
significantly damaged by plugging, even over the
well lifetime.

855
 5 WATER QUALITY REQUIREMENTS FOR FRACTURED INJECTION WELLS SPE 21439

FIELD PERFORMANCE OF VATER INJECTION VELLS

A period of injection without fine filtration from
The downhole water quality measurements discussed October 1989 to November 1990 has shown no
above, together with our increased understanding of reduction in injectivity from any of the 5 Forties
the role of thermal fractures in water injection Charlie wells.
prompted field trials of seawater injection with
and without fine filters. Following the success of fine filter by-passing on
Forties Charlie, the fine filters were removed from
Removal of Fine Filters in Forties Field the remaining platforms in August 1990. The Main
Sand formation and waterflood was sufficiently
Fine filters were recently removed from all water similar to the Charlie Sand, that similar measures
injection wells in Forties reservoir, following an were felt to be appropriate. Again, no reduction
initial trial period in which filtration was by- in injectivity was observed.
passed on the Forties Charlie platform, with no
consequent loss of injectivity. Filtration in Other Fields

The Forties formation is a soft, poorly In late 1990, following the experience in Forties,
consolidated sand with a Young's modulus of the water quality specifications applied to Ula
approximately 0.1x10 6 psi. This implies that it field have also been relaxed, leading to the by-
will respond to water injection in a somewhat passing of fine filters. Ula is a consolidated
different manner from the consolidated formations sandstone reservoir, in which field performance 10
discussed above. In particular, thermal stress shows clear evidence of thermal fracturing, which
effects are probably smaller than other stress substantially influences well performance. A
changes around the wellbore associated with the particularly striking feature is that injectivity
:aised fluid pressure. However, the rock strength has been shown to increase in response to
IS also very low and there is evidence from well reductions in the well-head injection temperature.
test analysis and hydraulic impedance testing
(HIT)14 for deformation of the sand around the Simulation suggested fracture lengths which are
wellbore, including increases in porosity and the dependent on local formation permeability and
existence of substantial cavities or fractures. injection rates and pressures, but which are
Mechanical damage including disaggregation is commonly in excess of 50 m. Inclusion of fines
demonstrated by the fact that most injection wells modelling in the fracture demonstrated that
have some degree of sandfill. HIT measurements on sediment loadings of up to 10mg/1 could be injected
three wells in the Forties Main Sand indicated into the Ula wells without harming injectivity.
fracture or cavity dimensions in the range from
metres to tens of metres in length, and perhaps one Initial indications from the field do not show any
metre in height. The effect both of disaggregation loss of injectivity attributable to by-passing of
and of fracture formation is to allow the the fine filters.
injection of poorer quality water.
It should also be observed that seawater injection
It was considered necessary first to conduct trials into certain North Sea fields has been carried out
on a limited part of the Forties injection system from the start in the absence of fine filters, and
before total removal of fine filters. There was n~ without any action being taken during bloom
prece~ent for rem~v~l of filters from an existing periods. It appears there has been no loss in
facilIty. In addItIon, there was some evidence for injectivity, and no requirement for backflowing of
effects from face-plugging in circumstances other the wells.
than straightforward seawater injection.
Injectivity was found on occasions to decrease
after a work-over, perhaps indicating plugging due
to dislodged corrosion products. Produced water CONCLUSIONS
injection, with a significantly higher
concentration of oil and solids than seawater, is 1. Measurements from the field demonstrate that
known to give rise to some injectivity problems. downhole water quality is often very much worse
In addition, bullheading of acid to the perforated than the water quality immediately downstream of
zones has on some occasions (but not others) been fine filters, and may sometimes be worse than raw
shown to increase injectivity. intake seawater. It appears that much of the
downhole particulate matter originates from
The Forties Charlie platform was selected for corrosion of the tubulars, which may be associated
initial injection trials in the absence of fine with imperfect control of oxygen levels. In
filters. Consideration was given at this stage to practice, fine filters are therefore ineffective in
the possibility of early water breakthrough which controlling downhole quality.
might occur if extensive horizontal fracture growth
was promoted by face-plugging. However, both the 2. Vater injection in many wells, including those
earlier HIT measurements in the Forties Main Sand, measured for water quality, has taken place over a
and subsequent measurements in the Charlie Sand period of years without injectivity losses due to
after removal of the fine filters, indicated no face-plugging. This is in spite of the fact that
fracturing in excess of 100 m from the wellbore, the downhole water quality lies much outside the
and therefore no effect on sweep. In confirmation, specifications set for fine filtration on the basis
pressure fall-off testing gave evidence of small of pore size measurement and linear core flooding.
positive or small negative skins, with no clear
evidence for extensive fractures, which can yield 3. The ability of wells to accept a relatively poor
highly negative skins. quality of water is explained by fracturing and
other forms of deformation caused to the reservoir

856
SPE 21439 P.J.CLIFFORD, D.W.MELLOR, T.J.JONES 7

rock by the injection process itself. In 7. Perkins, T.K. and Gonzalez, J.A., "Changes in
particular, wells in consolidated sandstone Earth Stresses Around a Vellbore Caused by Radially
formations, which inject water cooler than the Symmetric Pressure and Temperature Gradients," SPEJ
initial reservoir temperature, are very commonly Vol. 24, No.2, pp.129-140, April 1984.
fractured due to thermal stress effects.
Fracturing increases injectivity, and creates a 8. Morales, R.H., Abou-Sayed, A.S. and Jones,
large sand face area which takes a much longer time A.H., "Detection of Formation Fracture in a
to plug than the original perforation tunnels. In Vaterflood Experiment," paper SPE 13747, presented
some circumstances, fractures are capable of at the SPE Middle East Technical Conference and
extension to create new sandface, and maintain Exhibition, Bahrain 11-14 March 1985.
injectivity, in response to any plugging which
does occur. 9. Garon, A.M., Lin, C.Y. and Dunayevsky, V.A.,
"Simulation of Thermally Induced Vaterflood
4. By-passing of fine filters in seawater injection Fracturing in Prudhoe Bay," paper SPE 17417,
wells in North Sea fields has not resulted in any presented at the SPE California Regional Meeting,
detectable change in downhole water quality or loss Long Beach, 23-25 March 1988.
of injectivity over a period of months. Filtration
to a solids content lower than about 10 mgll is not 10. Prado, L.R., van Kruysdijk, C., Niko, H.,
now considered necessary for fields of that type. Sancevic, Z.A. and Rodriguez, R., "Falloff Testing
a Vaterflood-Induced Fractured VeIl in Vestern
5. It remains important to examine water quality Venezuela," paper SPE 18142, presented at 63rd SPE
issues on a field-by-field basis. Loss of Annual Technical Conference and Exhibition,
injectivity may result from injection of water with Houston, 2-5 October 1988.
high solids or oil loading into certain formations.
However, specifications for water quality still 11. Svendsen, A.P., Vright, M.S., Clifford, P.J.
need to give regard to possible fracturing. and Berry, P.J., "Thermally Induced Fracturing of
Ula Vater Injectors," paper SPE 20898, presented at
Europec 90, The Hague, 22-24 October 1990.
12. Clifford, P.J., Berry, P.J. and Gu, H.,
NOMENCLATURE "Modelling the Vertical Confinement of Injection
VeIl Thermal Fractures," paper SPE 20741, presented
A area of fracture face (m 2 ) at the SPE Annual Technical Conference and
C skin damage constant (dimensionless) Exhibition, New Orleans, 23-26 September 1990.
d sk skin thickness (m)
k sk skin permeability (md) 13. Fletcher, A.J.P., Lamb, S.P. and Clifford,
k rock rock permeability (md) P.J., "Factors Governing Formation Damage and
L cumulative flux (m 3 /m 2 ) Injectivity of Polymer," paper SPE 20243, presented
Psk pressure drop across skin (psi) at the SPE/DOE Seventh Symposium on Enhanced Oil
Q water injection rate (m 3 s- 1 ) Recovery, Tulsa, 22-25 April 1990.
\.I water viscosity (cp)
14. Holzhausen, G.R., "Impedance of Hydraulic
Fractures: its Measurement for Estimating Fracture
Closure Pressure and Dimensions," paper SPE 13892,
REFERENCES 1985.

1. Patton, C.C., "Vater Quality Control and Its

Importance in Vaterflooding Operations," JPT SI METRIC CONVERSION FACTORS
(September 1988) pp.1123-1126.
ft x 3.048 E-01 m
2. McCune, C.C., "On-Site Testing to Define md x 9.689 233 E-04 \.1m 2
Injection-Vater Quality Requirements," JPT (January psi x 6.894 757 E+03 Pa
1977) pp. 17-24.
3. Barkman, J.H. and Davidson, D.H., "Measuring
Vater Quality and Predicting VeIl Impairment," JPT
(July 1972) pp. 865-870.
4. Pautz, J.F., Croker, M.E. and Valton, C.G.,
"Relating Vater Quality and Formation Permeability
to Loss of Injectivity," paper SPE 18888, presented
at SPE Production Operations Symposium, Oklahoma
City, 13-14 March 1989.
5. Todd, A.C., Kumar, T. and Mohammadi, S., "The
Value and Analysis of Core-Based Vater-Quality
Experiments as Related to Vater Injection Schemes,"
SPEFE (June 1990) pp. 185-191.
6. EI-Hattab, M.1. , "GUPCO' s Experience in
Treating Gulf of Suez Seawater for Vaterflooding
the EI Morgan Oil Field," JPT (July 1982) 1449-
1460.

857
SPE 21 43 9

TABLE 1 TABLE 4

SUSPENDED SOLIDS IN ULA VATER INJECTION SYSTEM PARTICLE SIZE DISTRIBUTION IN FORTIES CHARLIE VATER
INJECTION SYSTEM AFTER REMOVAL OF FINE FILTERS
UPSTREAM DOVNSTREAH INJECTION DOVNHOLE DOVNHOLE
FINE FINE MANIFOLD A-OS A-03 Number of SVP U/S DIS VIR FC4-1 FC1-2
FILTERS FILTERS particles
larger than:
3.5mg/l 0.1mg/l 0.3mg/l 6.0mg/l 7.5mg/l
2lJm 3849 2577 2471 2341 4499 5323
4pm 565 397 402 298 570 886
TABLE 2
8pm 60 41 48 31 59 84
PARTICLE SIZE DISTRIBUTION IN ULA VATER INJECTION
SYSTEM 16pm 3 3 4 2 5 6

Number of U/S DIS VIR A-OS A-03 SVP - Sea winning pumps
particles U/S - Upstream of the fine filters
larger than: DIS - Downstream of the fine filters
VIH - Vater injection header
2pm 6653 58 555 17010 32517 FC4-1 - Downhole well FC4-1
FC1-2 - Downhole well A-03
4pm 900 2 25 3327 3874
8lJm 73 0 0 418 462
16lJm 5 0 0 54 56 TABLE 5
PERMEABILITY STRUCTURE OF LAYERED RESERVOIR MODEL
U/S - Upstream of the fine filters
DIS - Downstream of the fine filters
VIR - Vater injection header LAYER THICKNESS/ft PERHEABILITY/md
A-OS - Downhole well A-OS
A-03 - Downhole well A-03 1 50.2 2.0
2 26.7 7.5
3 13.3 7.5
4 40.0 34.7
TABLE 3 5 32.8 67.1
6 10.6 78.6
PARTICLE SIZE DISTRIBUTION IN FORTIES CHARLIE VATER 7 21.2 78.6
INJECTION SYSTEM BEFORE REMOVAL OF FINE FILTERS 8 29.9 30.1
9 20.3 4.9
Number of SVP U/S DIS VIR FC4-1 FC1-2
particles
larger than:
2pm 4299 4767 3842 3000 10726 6844
4lJm 466 524 423 238 1505 992
8lJm 38 36 31 19 73 62
16lJm 1 2 3 1 3 1

SVP - Sea winning pumps

U/S - Upstream of the fine filters
DIS - Downstream of the fine filters
VIH - Vater injection header
FC4-1 - Downhole well FC4-1
FC1-2 - Downhole well A-03

858
 SPE 21 43 9

BOOSTER
DEOXY/
COARSE PUMP SP3
VACUUM

~~~
. . . - POLVELECIROLYIE
ALTERS
TOWER
ANTIFOAM
~
OXYGEN SCAVENGER

WATER MANIFOLD
I - SP1

INLET
WATER

1.--,.+----11 RESIDENCE DRUM ~

t-SP2
BIOCIDE
ANEALTERS

SP1 UPSTREAM OF ANE ALTERS

SP2 DOWNSTREAM OF ANE ALTERS
SP3 INJEC110N MANIFOLD

Figure 1: ULA WATER INJECTION SYSTEM

8 Gravimetric analysis
Suspended solids
l!! 6

-
:!!
4

o
-- Upslnlam
fine fillers
Downstream
fine filters
;.;.:.;.;.:.:.;.:.:.:::;:::::::::.:::::::;:::::

Injection
manifold
Sample point
Downhole
ADS
Downhole
A03

Figure 2: SUSPENDED SOLIDS IN ULA INJECTION WATER

859
Sr-t: 2 1 43 9

40000 Particle analysis

Total counts in 500 microlitres
!30000
- 20000
~ 10000

o - - _ . : - : . : . ; . ; . ; . ; . ; . : - : . ; . : - : . ; . ; . ; . : . ; . ; . ; . ;.....

Upsteam
fine flhers
Downstream
fine fibers
Injection
manifold
Sample point
Downhole
A05
Downhole
A03

Figure 3: PARTICLE COUNTS PER O.5ml IN ULA INJECTION WATER

4 Total suspended solids

Pre removal fine filters
! 3
- 2
~

o Sea winning Upstream Downstream Downhole Downhole

pump fine fibers fine fihers 4-1 1-2
Sample point

Figure 4: SUSPENDED SOLIDS IN FORTIES CHARLIE INJECTION

WATER, BEFORE FINE FILTER REMOVAL

4 Total SuspendeCl solids

Post remoVal fihe filters
~ 3
2

......................:- ..;.;.:.;.:-:.:.;.;.
o Downstream Downhole Downhole
Sea winning Upstream
pump fine fibers fine fihers 4-1 1-2
Sample point

Figure 5: SUSPENDED SOLIDS IN FORTIES CHARLIE INJECTION

WATER, AFTER FINE FILTER REMOVAL

860
 8000. r ,- - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

~
g 123. w ,
a:
I-
U)
2i
60 days 200 days 300 days 600 days 700 days :l
lwZ
a:
7000. r:r/.!
~<\_ ......m-CC:::.'-L\
N 0. I: /, .

.i L'd
V ~\---~.::-::::-:-::l__--<:-
:.-::-::-::::.-=.--~:=:~::::'L
6000.
C = 0

C = 0.01

C = 0.1
5000. 1 1 1 1 1 I I 1 I
O. 100. 200. 300. 400. 500. 600. 700.
TIME (days)
X-DIST (It)
Figure 7: EFFECT OF WATER QUALITY
Figure 6: FRACTURE GROWTH WITHOUT FINES ON INJECTOR BOTTOMHOLE PRESSURE
OVER A PERIOD OF TWO YEARS (HIGH CONDUCTIVITY FRACTURE) (HIGH CONDUCTIVITY FRACTURE)

CD
!!1
246

g
I-
123. g
I-
U)

mllill
300
250
200
U) 2i ~
2i N
N

FRACTURE BOUNDARY

o 260
XDIST (l1)

X-DIST (It)
Figure 9: DISTRIBUTION OF SKIN DAMAGE
Figure 6: EFFECT OF WATER QUALITY
OVER FRACTURE FACE: 500 TO 700 DAYS, C = 0.1
ON FRACTURE SIZE AT 700 DAYS
(HIGH CONDUCTIVITY FRACTURE)
(HIGH CONDUCTIVITY FRACTURE)

SPE. 21 It 3 9
 8000.

III~
5 w .'"
/'
~ a: ./ . Ii .___ _.-----i
II> ::J
.__-j _/'/! /./.-- i -----.---.-j
q II>
II>
W
7000.
_ y-/ v'-- L---...---- i_.
N
a:
0..

6000.
00. 400. C = 0

C = 0.1

C = 1.0
5000. 1 1 ! ! 1 ! ! 1
O. 100. 200. 300_ 400. 500. 600. 700.
X-DIST (ft) TIME (days)
Figure 10: EFFECT OF WATER QUALITY Figure 11: EFFECT OF WATER QUALITY
ON FRACTURE SIZE AT 700 DAYS ON INJECTOR BOTTOMHOLE PRESSURE
(LOW CONDUCTIVITY FRACTURE)
(LOW CONDUCTIVITY FRACTURE)

!
245

g
~
II>

!iii
llillill
700
500
300
i5 tillm
N

FRACTURE BOUNDARY

o 260
X-DIST (ft)

Figure 12: DISTRIBUTION OF SKIN DAMAGE

OVER FRACTURE FACE: 500 TO 700 DAYS, C = 0.1
(LOW CONDUCTIVITY FRACTURE)

Spr. 21439