Society of Petroleum Engineers

SPE 25589

Development of the Ghawar Scale Squeeze Simulation Model:

Using Core Analysis To Evaluate Treatment Chemicals
J.D. Lynn, Saudi Aramco

Copyright 1993, Society of Petroleum Engineers, Inc.

This paper was prepared for presentation at the SPE Middle East Oil Technical Conference & Exhibition held in Bahrain, 3-6 April 1993.

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, 163245 SPEUT.

Section I, Abstract
Calcium Carbonate scaling problems have been detected in three of the
Aramco Southern area fields. As of 6/88, these fields have shown scaling
in more then 70 % of the wet producers. The scaling has resulted in loss
of production capability and created potential safety hazards throughout
the production string, and into the gathering areas. Various methods
are being evaluated in order to find a cost effective method of eliminating
or controlling the scale problem. The following methods have been
evaluated as possible alternatives to solve the problem:
. Tubing replacement
Physical or chemical removal of the scale (HQ)
Down hole chemical injection
Scale inhibitor squeeze
The method selected for application to the Arabian fields was the
precipitation inhibitor squeeze. Core analysis is playing a major role in
the development of the squeeze method, and in defining procedureswhich
will optimIZe the field treatment.
The final result of the testing was a core based production model which
could accurately simulate field behavior of inhibitor treated wells. The
data would then be applied to optimization of early field trials. The
generated data would then be applied to formulation of a mathematical
model for field application. This paper will illustrate the various
applications the core testing was used for, and show the relationship of
laboratory data to the formation data, and describe the use of the core
data in the optimization of the field treatment.
Section II, INTRODUCTION
Scale mitigation is a primary concern in the Ghawar field of Saudi Arabia.
Calcium carbonate scale is a factor in most wet producers, at greater then
70 % water production. The scale is believed to be generated through
up-hole degassing and loss of C02, reducing the natural solubility of the
Ca + + and HC03-. The material is generally precipitated in the upper
200-600 feet of the production tublers, and does not pose a threat to the
actual formation. The 'precipitation of the scale does cause major loss of
production through filhng of the production string, and down time related
to remedial activities. Precipitation of the scale in the downstream
assemblies caused environmental and safety concerns related to the
inability of valves and safety cutoffs to function properly.
Various remedial actions were investigated. Prior to 1986 chemical (HCL
wash) and mechanical (scrapers) were used to correct the problems, or
in very sever cases, tubing replacement was used. These methods were
labonntensive, and in the case ofchemical removal ofscale or replacement
of tubing were considered economically unacceptable. They were also
considered "holding actions" since they did not address the long term
mitigation of the problem.
After1986 investigationswere started into the use of"inhibitortreatments"
to prevent scale formation. These methods would be directed towards
preventing the precipitation ofscale rather then concentrating on removal
of the scale after the fact. The major feature required of the treatment
was long term economic feasibility. Secondary concerns were the
treatment should be non-damaging to the formation over several treat-
ments, and environmentally the material should be non-destructive. To
these means, two methods were deemed feasible. One was down-hole
injection of inhibitor using high pressure low volume pumps and capillary
tubing, and the second was a precipitation SC\ueeze. In both cases an
ATMP, Aminotri(methylene phosphonic aCid), type chemical was
evaluated. Early in the evaluation the down hole injection method was
found to be economically undesirable due the followmg problems:
A. One pump skid assembly would need to be positioned at each
subject well. Several hundred would need to be on-line at any
given time through-out the field.
B. Maintenance of sophisticated, low volume, high pressure
pumps under the harsh Arabian environment would be impos-
sible, without the construction of artificial environments at each
site.
C. Man-power and monitoring/scheduling requirements were
veryhigh, with rotational filling ofseveral hundred chemical tanks
a constant necessity.
D. Long term stability of the chemical stored in the on-site tanks
was questioned.
E. Power requirements of the pumps was a problem.
F. Failure of the pumps would result in precipitation of the scale
restarting.
The inhibitor squeez~ tpethod had been used in ~ral applications in the
Prudhoe Bay regionLl] and in the North seaL2], very successfully, but
treatment life has been short. The economic factors in the Arabian fields
required a treatment life of at least two years for each treatment. The

587
2 DEVELOPMENfOFlHE GHAWAR SCALE SQUEEZE MODEL: USING CORE ANALYSIS TO EVALUATE
TREATMENT CHEMICALS
main thrust of the core based assessments was to understand the
relationship between the injected chemical and precipitation of the
inhibitor in the formation, and examine the controlling mechanisms of
the precipitated chemical solubility in the formation brine during pro-
duction. These data would first be applied to optimization of the field
squeeze method, then used in the generation of a mathematical chemical
squeeze simulator to study longer term effects.
Section III, EARLY INVESTIGATIONS
A. Assumptions
Various assumptions had to be made concerning the application of the
chemical in the field. For the model calculations we have assumed two
sections of the reservoir (oil and water) and that the two zones have the
following properties, and inter-relationships:
1. WATER ZONE
a. The water zone is contiguous.
b. The water is being produced from this single zone.
c. The water zone is a watered out section of the oil producing
zone and is at irreducible oil saturation. Effective porosity must
be determined to establish treatment volumes.
d. The water zone is homogeneous throughout.
e. The water zone has an absolute porosity of 18 %.
2. OIL ZONE
a. The oil zone is contiguous.
b. The oil is being produced from this single zone.
c. The oil zone overlies the water zone.
d. The oil zone is homogeneous throughout.
e. The zone has an absolute porosity of 18 %.
f.· Because of the differences in capillarity between the oil and
water zone, the water based chemIcal will only enter the zone
producing water.
Although admittedly simplistic, these assumptions led to a suitable
understanding of the basic chemical/formation behavior when properly
applied, and analyzed in the proper prospective.
B. Theoretical considerations·
We can generally break the squeeze process into a relative series of
functions, which, if taken to~etherwillprovide a mathematical simulation
of the interaction of the vanous components of the inhibitor/formation
interaction. The equations are as follows:
I b - f(T ,) f(l ,) ( I)
The inhibitor concentration in the produced brine, Ib' is a function of
the contact time, T c, and the solubility of the inhibitor in the produced
water, Is.
T, - f(FV)b f(SA)/ (2)
The contact time, Teo is a function of the velocity of the formation brine,
FVb' and the available surface area of the preCIpitated chemical, SAl.
SA/ - f(l ,)f(•• )f(a)f(H) (3)
The available chemical surface area is a function of the injected chemical
concentration, Ie, the effective porosity, 4>. , pore tortuosityand geometry,
a, and the formation heterogeneity, H.
FV b - f(D)f(H) f(P b ) f(4).) ( 4)
The velocity of the fluid moving through the reservoir, FVb' is a function
of the distance from the well bore, D, the thickness of the reservoir, H,
the volume of produced water, Pb, and the effective porosity, 4>.. And
[mally,
4>.-4>,xS (5)
Sw is the water
4>. is the effective porosity, 4> I is the total porosity and
saturation of the formation, as a fraction of the total porosity.
All of these functions can be examined and quantified in the laboratory,
but it would be prudent to examine only those which we can easilyestablish
control over in the field operations.
588
SPE25589
From equation 1, we can see that we can either adjust the contact time,
or the solubility of the chemical in the brine. Without complete
re-formulation of the injected chemical and extensive advanced chemical
synthesis methods we will have little hope of changing the solubility of the
chemical. This means we should concentrate attention to the modification
of the contact time.
From equation 2, we can see that the contact time is basically a function
of two components, the produced fluid velocity, and the surface area of
the chemical. The surface area function can be broken down in the
functions presented in equation 3. Of these four functions, the only one
which can be controllable is the initial concentration of the injected
chemical. This process was investigated during the core modeling
experiments (sim A), by reduction of the chemical concentration in the
slug by 50 %, while doubling the total volume of the slug. In other words
the same amount of chemical was injected, but in a more dilute form,
overa largersection of the reservoir. While this reduction in concentration
produced better overall returns, the general pattern of high initial returns,
with high returns through the process remains. This means wasted
chemical, and shortened chemical life.
The second function of equation 2 relating to the fluid velocity is broken
down to four functions in equation 4. Of these, only the function relating
to the distance from the well bore can be considered controllable. Based
on the flow data collected during the current simulation, distance from
the well bore appears to have the greatest promise of delivering a
consistently low return over an extended period of time. Figure 8 shows
calCJJlated dissolution rates for the core simulations. The trends show that
as the flow rate increases the dissolution rate decreases. The basic
inference of this is that if the flow rate is high enough, . allowing only
enough time for a minimal amount of chemical to be dissolved, the life
of the treatment can be greatly enhanced. It is probable that some
combination ofchanges in distance (decreased overflush), and in chemical
band width (chemical volume) can be found which will provide chemical
concentrations in the range shown to be effective in preventing scale.
Adjusting these rates to achieve only the concentration of chemical
required for mitigation is critical to extending the treatment life.
C, Treatment phases
A major concern was the reduction of the total volume of the treatment
to a minimum for two reasons. One, the major costs of the squeeze
treatments tended not to be the actual chemicals, but cha~es involved
with transportinjr; and pumping large volumes of water mto remote
locations. Pumpmg charges average $20.00/bbl. Two, since many of the
wells were alreadywet producers (usually> 70 % water CJJt), the injection
oflarge volumes ofwater made returning them back to production a major
operation. This increased the cost of the jobs by requIring N2 kick-off to
restart the wells. The vendor's recommendation, for a 14-16 foot
treatment radius was considered as a starting point of the studies.
The field inhibitor squeeze treatment as adopted by Aramco consisted of
four phases of injection into the producing formation. The first is a
pre-flush or spearhead. The spearhead functioned to remove hi2IJ
concentrations of divalent cations from the vicinity of the well bore. 'fo
achieve as low a pre-flush volume as possible, the calculated volume was
tied to the total concentration of these cations in the pt?duced formation
brine. This was followed by the chemical slug, consISting of the ATMP
diluted with a low salinity aquiferwater, a spacer, and finally an overflush
to position the band of chemical a fixed distance from the well bore. The
volume of ATMP used was calCJJlated to be 1% of the daily water
production of the treated well. This value was based on the vendor's
recommendations. The largest and most critical volume examined was
the overflush volume. Optimization of this parameterwould give the most
cost savings to the operation. The effects of each of these components
was evaluated using reservoir core material.
D, Chemical/Core/Formation Brine Interactions
First, and most critical to the evaluation of ATMP is the understanding
that the chemical selected for injection is a multi-protic acid, of
considerable strength. The dissociation reactions and reaction with the
carbonate formation is shown in figure 1. Second, the anionic ~es
developed will react not onlywith the Ca + + generated in situ, But Ca ++
and Mg+ + species present in the formation water. Third, and, again
critical is the understanding that chemical returns will be a function of the
solubilityof the chemical in the produced formation brine. This solubility
is in tum a function of the contact time between that formation brine and
the chemical precipitated in the formation.
SPE25589
1. Pre-Flush Requirements
The pre-flush provides two functions. One, it prepares the well bore for
chemical injection. Two, since the chemical has been shown to be reactive
with any Ca + + species, including that in the formation brine, all of the
extra material must be swept from the region of the well bore out to the
projected radius of treatment by a fluid low in total Ca and any other
divalent cations. Figure 2 shows a plot of time fO start of phosphonate
precipitation as a function Ca + + concentration 31 As can be seen, to
allow for a total of 30 minutes pumping time, the total available Ca + +
ions !l1~st ~e bel?W1000ppm. Levels higher than this will cause pre-mature
prt;ClpltatlOn, WIth the danger of near well bore damage through plugging.
ThIs plot can be used to derive the thresh hold level of Ca + + permissible
in the formation water for a given time of inhibitor injection.
To evaluate the volume of preflush required to produced the appropriate
level of Ca + +, a series of core floods under reservoir conditions
(wettability restored, reservoir and pore pressures) was used to evaluate
the efficiency of brine/brine displacement in the treatment formation.
The core was at irreducible oil saturation. The water used for the Ca + +
displacements was the aquifer water to be used for preparation of the
spearhead on-site. Injection rates were calculated based on probable field
rates during the treatment operation. The field rates were scaled to the
core rates using calculated interstitial rates (frontal velocities). Each of
the cores was saturated with a Ca solution of known concentration, and
the total volume of fluid need to displace the Ca + + from the core to a
specified level was determined. Injected volumes were calculated as a
function of the pore volume of the samples. The resultant plots are shown
in figure 3. Using figure 3, the initial concentration of Ca + + in the
formation brine, and the required thresh hold to provide the time required
for injection (from ftgure 2), the required preflush volume can be
interpolated based on the total volume of overflush injection (assuming
1overflush volume = 1 pore volume). Figure 2 is a function ofthe chemical
used for treatment, and will varywith the type of chemical, concentration
of chemical injected, and total 'IDS of divalent cations present in the
brine. Figure 3 is function of the core tortuosity, wettability, and residual
oil saturation for the subject formation. This plot will provtde a minimum
pre-flush required for the injection. Consideration will have to made for
the actual rate of Ca + + generation in situ, by the acid.
2. Effects of Lithology/Reaction time
Since this is an acid, reaction rates will be strongly dependent on
lithological character. The zones being treated in this work are mainly
limestone, but can in some areas be 100 % dolomitized. The required
shut-in time for the treatment will be controlled bythe rate of the reaction.
We must optimize this shut-in time to allow sufficient time for complete
reaction of the chemical, but not waste production time with overly long
shut-ins.
Several plugs were selected from the subject reservoir, each with varying
amounts of dolomite present. The plugs with similar composition were
stacked to provide 6 mch in length by 15 inch in diameter composites.
Since we were dealingwith an acidreaction, care was used to insure proper
reservoir pressure (confining and pore pressures) and temperatures were
used for the tests. tn the case of this test we were looking at a "worst case"
scenario since the cores had only been saturated with formation brine,
and were not at irreducible oil saturation. This would leave more of the
core exposed to potential damage.
One effective pore volume of the field strength chemical inhibitor was
injected into the cores, and then shut-in for time periods ranging from 24
to 96 hours. Formation water was then injected through the cores, and
chemical returns as a function of throughput and Ph were monitored. It
was important to note that the pH had to be measured on brine which had
been de-JtlISSed to remove the dissoloved C02, which would contribute
to low pll values. Four additional composites were injected with several
pore volumes of the acid, to examine the long term effects on the near
well bore. The results of these tests are shown in figures 4 and 5. The
return pH can be considered a reflectance of completeness of reaction.
The plot shown in figure 4 (for 24 hours shut-in) ispresented as an example.
One day was insufficient as shown by the acid pH numbers during the
early returns. Neutral returns were noted after day 2. The shut-in was
therefore set at a minimum of 48 hours. It was also noted that even after
96 hou~ the reaction in a pure dolomite core did not appear to go to
completion.
589
J. D. LYNN 3
As shown in figure 5, long term effects on the near well bore can be
considerable, and in the case of a pure limestone core, resulted in matrix
failu~. This led us to believe that an acid activity retarder would be
requIred to protect the well bore, and facilitate precipitation of the
chemical at the proper distance from the well bore. Protection of the near
well bore will increase in importance as the volume of inhibitor is increased
and as the frequency of retreatment is increased.
3. Inhibitor Reactivity Retarders
Since near ,!e.1I bore damage was ~hown to be a potential problem, the
use ofan actMty retarderwas exammed to slow the rate of reaction. Three
different chemicals were examined, again using core material, at reservoir
conditions. Brine permeability was measured on each of the samples.
Four concentrations of retarder were examined: 0.05 %, 1.0 %, 2.0 %
and,3.0%.
Mter the base permeability was determined, injection of the acid and
retarder mixture into the core was started. 100 pore volumes of the fluid
was pumped through the core sample. A chart recorder was used to
momtor the pressures as the retarded acid was injected. Following acid
injection, bnne permeability was again determined. Graphical repre-
sentation of the data is presented in Figure 06.
At concentrations below 3.0 % none of the retarders were effective in
protecting the formation. RetarderA gave the least protection, and should
not be considered for this application. Based on these tests the retarder
C gave the best protection at the 3 % level, although even at this level
the return permeability to the brine was increased by 100 %.
In conjunction with this testing a problem was noted concerning the
formation of an emulsion when producing retarder A and oil. It was not
known if this emulsion was a relic of the testing procedure or a possible
problem resulting from actual retarder/crude 011 interactions. Further
study of the problem indicated that the formed emulsion would break in
less then the expec;t"d shut in time for the field treatment, and should
cause no problemsL4J.
The acid retarders examined in the study were found to be ineffective at
the field strengths being used. Use of these chemicals provided no
perceived benefit. The alternatives are as follows:
1. Increase the retarder concentration until it becomes effective
2. Discontinue use of any retarder during the treatments
3. Locate a more robust retarder
Option 1 may cause problems due to poor mixing properties of the
retarders at high concentration. The retarders have problems with the
formation of an insoluble material above 4 % concentration. The study
indicated that a concentration of 8 % was required before the material
was effective.
Option 2 may result in damage to the well bore environment, and possible
collapse after several treatments.
Option 3 is being evaluated, but with little success at present. Small
quantities of the retarder continue to be added to the treatment, but the
benefit derived from them is consider trivial.
Section IV. RESERVOIR MODEL CONSTRUCTION
The preliminary data presented above was then used to construct a core
based model of the squeeze procedure. A three foot, reservoir condition,
model was used to simulate the formation. Core was taken from the subject
formation and prepared as follows.
A. CORE PREPARATION
Extracted, 15" in diameter plugs from the subject formation were used
for the testing. Extracting of core samples generally results in destruction
of the natural wettability of the samples. In order to get a more
representative simulation of the reservoir, it was decided to "restore" the
wettability of the core samples by the following method:
1. The samples were saturated with the simulated formation
brine, and complete saturation.
2. Theywere then placed under the saturating brine, in the sealed
container for four weeks at the reservoir temperature of 95 ·C.
3. The samples were then dynamically flushed to Swir with dead,
filtered, formation crude oil.
4. A second four week aging under crude oil in a sealed container
at reservoir temperature.
4 DEVELOPMENT OF TIlE GHAWAR SCALE SQUEEZE MODEL: USING CORE ANALYSIS TO EVALUATE
TREATMENT CHEMICALS
The ends of the core samples were machined, and butted end to end. The
composite core was placed into a mounting jig and wrapped in two layers
of Teflon, followed by a 1/8" Viton outer sleeve. The total length of the
composites ranged between 55 and 85 cm. The apparatus constructed for
the study is shown in figure 7. Confining stress was placed at 2950 psi,
with a pore pressure of 1000 psig. The oven temperature was set at 95°C.
The apparatus was given 24 hours to reach stability. Oil flow was then
estabhshed in the production direction. Pressure drop was monitored,
and oil permeability at Swir was determined.
Mterdetermination ofoil permeability, the core was flushed to irreducible
oil saturation, at a high flood rate. This flush was also carried out in the
production direction using the simulated formation brine. The flush was
allowed to continue until no oil production was noted for a 24 hour period.
Final oil saturation at this point was calculated. Brine permeability was
determined at the irreducible oil saturation. The fluid permeability was
measured in the forward and reverse direction to insure that there would
be no problem from mobile fine material in the pore space. These
measurements were found to agree within 1% of each other. The core
was now considered to be in a state similar to a water producing, swe{lt
oil zone in the reservoir, the expected target of the squeeze treatment In
the field.
B. Pore Volume Determination
The pore volume which will be affected by the chemical will be only that
portion which contains water, and that portion which due to capillarity
will allow entry of the inhibitor into contact with the rock surface. In the
case of a strongly oil wet core, or in a sample with a highly bi-modal pore
size distribution, effective porosity contacted by the chemical may be only
a fraction of total porosity. Determination of this effective porosity was
conducted on the composite cores by the following method.
The core was at irreducible oil saturation, with formation brine as the
major phase. A brine containin~ 10,000 ppm KCl (5000 ppm k +) was
prepared. This fluid was then injected into the core at a rate consIStent
with proposed field injection rates. Incremental samples were taken every
5 cc ofproduction and analyzed by Atomic Adsorption for total K +. Based
on the break-through of the K + brine, the effective pore volume could
be calculated. In the case of this formation, it was found to be 10 % of
the total porosity. The core was then flushed with the formation bryne
until the remaining K + concentration was removed. The graphical
representation of these data is shown in figure 8.
It was noted that this was less then the expected volume. The calculated
values should have been the difference of the total pore volume and the
irreducible oil saturation, or the water saturation of the zone. The
remaining bypassed porosity appears to be from low capillarity portions
of the core material. This was later confirmed by mercury injection
capillary pressure data which showed a bi-modal pore distribution. The
volume of this bypassed material will be a function of down-hole injection
pressure. It is thought to pose only a minor influence on the field squeeze.
C. Injection procedures
The core was first injected with the spearhead/surfactant mixture. The
inhibitor/retarder mixture was placed into a sample loop, and injected
into the core at the proposed field injection rate. These rates were scaled
to frontal advance velocities (interstitial velocities) using the following
equations:
(2834xQ,)
FV· (6)
(Ax'!>,)
WHERE,
FV = FRONTAL VELOCITY, FEET/DAY
2834 = CONVERSION FACfOR
A = AREA, SQ CM
Or = RESERVOIR FLUID FLOW RATE, CC/SEC
<1>, = FRACfIONAL, EFFECfIVEPOROSITY
The area used for the calculations is taken to be the lateral surface area
of a right cylinder of height equal to the treatment thickness, and radius
equal to the treatment placement distance.
590
SPE 25589
Based on the area and porosity of the core sample used a rate, the frontal
velocity can be scaled from a radial system to the linear model system,
using the following equation:
(FVxAx'!>,)
Q,. 2834. . (7)
Where,
01 = LINEAR CORE FLOW RATE, CC/SEC
A = AREA OF TIlE CORE SAMPLE, SQ CM
FV = CALCUlATED RADIAL FRONTAL VELOCITY, FEET/DAY
<1>, = EFFECfIVEPOROSITY, FRACfIONALPOROSITYOFTIlE
CORE
Mter the chemical slug, the overflush and diesel cap were injected. The
core was shut in for 48 hours to insure complete reaction of the chemical
with the formation.
Based on the simulated treatment volume and the effective pore volume
of the composite core, A rate was determined which would match the
calculated frontal velocity at the proposed treatment radius. A series of
three different injection radii were investigated during the experiments,
one at 15 feet (sim A), one at 8 feet (sim B) and one at 6 feet (sim C).
In addition, Sim A used a reduced chemical concentration.
If the actual frontal velocity of the fluids is calculated, it is found to be
37.46 feet/day 15 feet from the well bore as opposed to 69.21 feet/day for
the 8 foot simulation. Since it has been shown that the contact time should
be a function of fluid velocity and chemical surface area, an instantaneous
point velocity has been used to scale the model rates to the well rates. The
frontal velocity at a point in the reservoir will be given as previously
calculated for the injection rates.
Following injection and shut in, the core was then placed back into
productionwith the simulated formation brine, at the appropriately scaled
rate, and incremental samples taken and sent to the chemistry unit for
total phosphonate analysis.
Each flood was taken to between 450 to 700 simulated production days.
The resultant data are presented in figure 9 (inhibitor concentration vs
simulated production days) and figure 10 (% total ATMP returns vs
simulated production days.
D. Discussion of results
Sim A showed the overall worst performance of the presented tests. This
is due to its greater radius (15 feet), and also apparently due the lower
concentration of the ATMP in the treatment slug. Overall returns were
substantially higher than either sim B or sim C. Field studies have shown
that the chemical is effective in concentrations as low as 5 mgfL, based
on no scale formation through visual assessment;]!froducing equipm~nt.
If this is case, then return levels in excess of 1 m are wasted chemical,
and will be reflected as shorter treatment Ii e. Indeed, in severaJ
treatments, there have been cases of thin phosphonate scaleprecipitation
following the well treatment. Sim B, and slm C were very simJlar in nature.
Both showed lower returns over the life of the treatment, although still
substantially over that required for well protection. Total returns were
approximately 15 % lower in the sim Band C cases, based on the returns
calculated at 400 simulated days. Sim C had used approximately4.8 % less
chemical by the 10 day mark then did sim B, but they were at virtually
identical levels by day 200, and finally by day 450, sim C had used
approximately 35 % less chemical, to achieve the same level of returns.
In all, the level of difference between sim Band sim C are close enough
to the calculated level of precision between any two simulation runs (+ 1-
4.0 %, mainly due to lithology, and precision oflowconcentration inhibitor
measurements) to make them statistically insignificant.
Sim A and sim C show very similar return curve characteristics in figure
10, while sim B seems unusual. AJso in figure 9 there appears to have
been an unusually high volume of chemical returned early In the run .(at
0.1 simulated days). This implied the reaction had not gone to completion
with the formation materiaJ in sim B. X-ray diffraction analysis showed
there was 4-8 % dolomite in the sim B composite, while the other two
showed no dolomite. This could be the reason for the inconsistent %
returns data. The test did serve to highlight the necessity of lithological
control within the treatment interval and the test samples. However, as
SPE25589
will b~ shOW!! the in the section on economic impact, the fact that the 6
foot simulation performed at least as well as the 8 foot simulation will be
the relevant point.
The solubility of the chemical must also be limited by the spatial
dist~bution of the matet!al following precipitation in the formation.
Studies conducted by Lewis, et.al (personal comm, 1992) indicate that if
the measured solubility of the inhibitor chemical in the formation water
the volume of chemical injected, and the calculated brine production i~
the wel~s is examined the material should provide no more then two weeks
protection. To date the longest treatment life seen in field operations is
m excess of 1000 production days.
This p?ssibly re~ec~ th~ effect from heteroge~eous precipitation in the
formation or a "dlstnbutlon effect". If the matenal was evenly distributed
then the actual water/chemical contact would be continuous in the po~
~te.m. It appears ~s t~ou~ t.he material, ~eing less then evenly
dlstnbuted, IS expenencmg lImited, not contmual, water/chemical
contact, resulting!n solubility being a function of the velocity of the fluid
through the chemical band zone and the distribution of the chemical in
the formation.
The indications point towards the flow channel size and the total amount
of "intra-granular" porosity as having a controlling influence on the
squeeze treatment life. The effects caused by geometrical and tortuosity
(rock Iithofaci~s ~nd petrofabric) factos in the ~ore samples should not
be overlooked m Importance. The use of formation core material for the
ini~ia! che~i~al effect model ~II enable these effects to be quantified.
This IS a cntlcal piece of data m later construction of the mathematical
simulator.
E. Cost impact of the optimization
To hi2hlight the impact of the optimization suggestions taken from the
core data, a case study was developed, around the following postulated
well conditions: treated thickness (wet) = 150 feet water rate = 9600
bbls/day, effective porosity 10 %, treatment concentration, 20% ATMP
chemical volume = 1 % of daily water production. Table 1 shows ~
comparison of the volumes of fluids required for the various squeeze
treatment simulated by this study
Using the parameters optimized based on the core model, total treatment
volume can be effectivelyreduced to 1/3 the originalvendor recommended
procedure. Average pumping cost per barrel is approximately $20.00.
Assumingall otherprices constant, moving from the vendor recommended
15 feet to an effective 6 foot radius will result in a 52 % decrease in the
cost of treatment of the test casewell. In addition, the reduction in volumes
allows the well to be put back into production quicker, and with less
returned water to dispose of.
SectioQ y. CONCLUSIONS
The test results reflected the predictions of the theorypreviouslrdiscussed
regarding the moving of the contact zone closer to the wei bore and
regulating the contact time, as a result in the change of fluid velocity
through the chemical precipitation band. This goes against the conven-
tional wisdom of typical squeeze applications, in which larger treatment
volumes and radii have resulted in extensions to treatment life. The reason
for this can be found as the result of the fundamental difference between
this method and standard squeeze techniques.
The method employed to deliver chemical to the treatment zone is a
precipitation reaction, catalyzed by the acid dissolution of the carbonate
formation and liberation of free Ca for precipitation, and resulting in the
precipitation of a modified chemical species.
In most adsorption / desorption treatments there is no modification of
the product, and basicallywhat you put in is what you get back, with some
loss to active sites in the formation. Chemical returns are then loosely
c~ntrolled b>: capillary retention, and weak molecular bonding to some
mmeral speCies. Larger volume would therefore mean extended time
before retreatment. Treatment life is a function of treatment volume.
We h~~ ~hown that the relationship of the precipitated chemical to its
solubilIty IS a complex function of chemical properties, interstitial fluid
veloc~ty, . and litholo~cally induced dis.tri~ution of the precipitated
chemical. m the formation. In order to optimIZe the treatment for a given
field envuonment numerous core based tests must be employed to obtain
characterization data on the chemical/formation relationship. The most
important are as follows.
591
J.D. LYNN 5
a. Inhibitor chemistry and reaction in the formation
b. Formation fluid/chemical interactions (brine induced pre-
cipitation, formation of emulsions)
~. ~otential .formatio~ damage from chemical activity (precip-
Itation plUgging, matnx damage, fines generation, )
d. Chemical additive requirements (retarders, surfactants,
demulsifiers)
e. Required time of reactions in the formation (shut-in times)
f. Potential competitive reactions in the formation (divalent
precipitation)
g. Lithology driven interactions (chemical distribution)
h. Pre-flush volumes
i. Placement distances
j. Potential treatment life and controlling factors
The core based model has the following advantages:
1. The constructed core based model allows for a quick and
affordable method of optimization for field treatments prior to
formulation of the appropriate mathematical models.
2. Potential damage from untried methods or treatment chemi-
cals can be easily evaluated under laboratory conditions without
the danger of well loss of formation damage. '
3..Economic optimization of treatments can be made and
numerous scenarios evaluated. '
4. Treatment can be broken in to their individual stages, and
examined on a one by one basis, determining ideal parameters
for each stage without interferences from other operating
parameters.
5. Treatment costs can be reduced by 50%, with no loss of
treatment life. This cost reduction is from reduced volumes of
fuids injected, saving pumping charges.
Under ideal situations, the data generated from the optimization testing
can then be used for mathematical simulation of the process and the core
model used for final verification of predicted effects based on the
simulator.
Acknowledgements
Thanks are given to the Saudi Arabian Ministry of Petroleum and
Minerals, and Saudi Aramco for permission to publish this paper. The
author further wishes to thank the professional, technical, and admin-
istrative staff at the Saudi Aramco Laboratories Department for their
support.
References Cited
1. Meyers, K.O., Skillman, H.L., and Herring, G.D., "Control ofFormation
Damage at Prudhoe Bay Alaska, by Inhibitor Scale Squeeze Treatment",
JPT, June 1985.
'*
Underdown, D.R. and Newhouse D.P., "Evaluation of Calcium Carbonate
Scale Inhibitors for Prudhoe Bay, Alaska", SPE ~per 15658, 61st SPE
annual Technical Conference, New Orleans LoUisiana, 1986.
2. Schuler, PJ., "Prevention of Down-hole Scale Deposition in the Ninian
Field" SPE Production Engineering, May, 1991.
'*
3. Lewis, AL., "ATMP Precipitation Tendencies", Saudi Aramco
Laboratories Department Report CSU-107/92, Aug. 1992, Dhahran,
Saudi Arabia.
4. Fritz, J.E., "Crude OilJWater Emulsion Stability Evaluations", Saudi
Aramco Laboratories Department Report EOLD-'15.22.10f87(a), March
1987, Dhahran, Saudi Arabia '
 DEVELOPMENT OF THE GHAWAR SCALE SQUEEZE MODEL: USING COAE ANALYSIS TO EVALUATE
6 SPE25589
TREATMENT CHEMICALS

o
II
H CH -. P -OH
oI / 2 I
0
OH- P-CH - N H
" 2 "" H
o "CH"2 -?P -OH
II
o
Figure 1: ATMP chemical structure

JO,OGO 1---I---;::;::::;:;:;:::::l;;;::==::::::;~I----T----1
a.ooo ~ -::._ Co

f
II

I,OlIO

IOOL....------'-------'-------'-----_-'---------'
o 10
,. ......................
. 10

Figure 2: Pr.eclpatatlon time as a function calcium concentration

1.4 - , - - - - - - - - , - ' - - - - - - - - , - - - - - - - - - , - - - - - - - - - - , - - - - - - - - ,

1.2

500 PPM 1000 PPM

0.'
2000 PPM

~oo..O. P..P!\lt .
0.'
1500 PPM
0.4 2500 PPM

0.2

o 1,000 10,000 11,000 20,000 H,lIOlI

INITW- c. CONCENTRATION, PPM

Figure 3: Initial calcium concentration vs. pre-flush volume as a function of

critical calcium concentration

592
SPE255U J. D. LYNN 7

8r----------,--------__r_--------.-----------,--.
30
7 t- . •
\ ~jt 4
• I-t~-- - - - - - - - - + - - - - - - - -..-.. /+-::-:~-:.-::~-:::.-?-~:~-~:.:.-:~"!"--:~f-:
:~~::;..~.•J-- ------l - - - ; - - f--+---l

1 .' ~ ~ - ~.... ......•..... f--+---l

i·-\·~~ •••........ _~_

t ..

4-\.... . ::::;~
22
. . . . . . - ..-e-..- .

3 ~..\ .~~~2 . m

2~~:~·

1
o 50 100 150 200
Through-put, treatment volume.

Figure 4: pH as a function of throughput, 24 hour shut·ln

~~~ ~~~
100,000 r--__r_1~0~0:,%~L~Sr-_.rO~50~1c;::OO=-.:1.:;:50:.=:200::r=250=;:=300:;-__r___i1~0~%:;:..=D~OrL=_r__,;_0~5Or-1~oor1~5O~2OOT~25Or3OO=;

. . . !~cj=t==1. . .

100 ···1· ···1·· . 1··········1·······1-/··········1············

10 j. + j·················f··········· I···· .. t + l · t

o 50 100 150 200 100% DOL 0 50 100 150 200 250 300 35% DOL
T-PUT, PV T-PUT, PV

Figure 5: Acid effects on rock matrix

593
 DEVELOPMENT OF THE GHAWAR SCALE SQUEEZE MODEL: USING CORE ANALYSIS TO EVALUATE
8 SPE25589
TREATMENT CHEMICALS

RETARDERB RETARDERC
10,000

1,000

100

10

1
0.5 1 2 3 0.5 1 2 3 0.5 1 2 3
'" RETARDER

Figure 5: Retarder effectiveness

6,000 I I

5,000 ~• .. - EFFECTIVE PORE VOLUME

~
t 4,000 1\
~
i
5
3,000 \
~
(J
+ 2,000
\ ...........

\
~

1,000 ~ ... ........

'-,
r'----
o
o 100 200 300 400 500
THROUGH-PUT, co

Figure 8: Pore volume determination

594
 tn
"'a
m
~

Figure 7: Expermental apparatus i

chart recorder
A A A A
~oOo
I !
, ii i
J
.. !
a:

I &1 t
'-
;D::-

~
100 PSIG (;; 1t : 1
L_.. .._.._.._.__.. _.l-. .._.._.._.. ._..l !-
P
~
500PSIG
01
co ~ z
01 SLEEVE PRESSURE = 2llIiO
!
1._••_-_.•_ .._ .._ .._ ..•.._ .. - ......_ .._ .._ .. _-_.._-._ .. _-_..__ .,j

~
~

;D::-~

f't"'~
.. [Boo PSIG BACKPRESSURE I
~~TOR
' :

I I
::~
~

r I 'I %

~,= =
OVEN TEMP = 95 "C)

'111 ~ BACKPRESSUR~ ~
1

SOURCE I
CD
 DEVELOPMENT OF THE GHAWAR SCALE SQUEEZE MODEL: USING CORE ANALYSIS TO EVALUATE
10 SPE 25589
TREATMENT CHEMICALS

SCALE SQUEEZE SIMULATION

ATMP RETURNS VS SIMULATED PRODUCTION DAYS

100,000 r--r--------r------~------....,...._-----___,

SIM A 12 FOOT CASE

eL 10,000 , .................................... SIM B 8 FOOT CASE

! SIM C 6 FOOT CASE

i "-
1,000 .' ..........•.l'l\;: ........ ......
" ,-'. ...-.....
~
100 ............................................................"
e :.~~.~.~ ~ ~ ~ ~.~.~.7:-:
.. .. .. .. .;:-:.~ .
10

0.1
0.1 1 10 100 1,000
SIMULATED PRODUCTION DAYS

SIM A SIM B SIM C

10 % CHEMICAL 20 % CHEMICAL 20 % CHEMICAL

Figure 9: ATMP returns vs simulated production

g60 r - - r - - - - - - - , - - - - - - - - - - r - - - - - - - - r - - - - - - - - - ,
I . SIM A 12 FOOT CASE
0 50 SIM B 8 FOOT CASE

1
~ 30
40
SIM C 6 FOOT CASE

. . . . . . . . . . . .......;.:~:;;.~,<~ .
'it.
uS 20
.... -;;...
.,.~~ .
,
110
eL •••••••••••••••••••••••••••••••••••••••
.... .
!O===:..:.....----...I....---------I...------....L....---------'
0.1 1 10 100 1,000
SIMULATED PRODUCTION DAYS

~MA ~MB ~MC

10 % CHEMICAL 20 % CHEMICAL 20 % CHEMICAL

Figure 10: % total ATMP recovered

596