SPE

SPE 23814

Kinetics of BaS04 Crystal Growth and Effect in Formation Damage

R.M.S. Wat, * K.S. Sorbie, * A.C. Todd, * Ping Chen, and Ping Jiang, Heriot-Watt U.
'SPE Members

Copyright 1992, Society of Petroleum Engineers, Inc.

This paper was prepared for presentation at the SPE Inti. Symposium On formation Damage Control held in Lafayette, louisiana, February 26-27, 1992.

This paper was selected for presentation by an SPE Program Commillas following review of informatioil 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 subJact to correction by the author(s). The material, as presented does not necessarily refleci
any position of til? Society of ~etroleum Engineers! its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society
of Petroleum Engineers. PannlSSlOllto copy is restricted to an abstract of not more than 300 words. illustrations may not be copied. The abstract should contain conspicuous acknowledg-
ment of where and by whom the paper Is presented. Wrne Ubrarian, SPE, P.O. Box 833836, Richardson, TX 75083-3838 U.S.A. Telex, 730989 SPEDAL.

ABSTRACT INTRODUCTION

In the North Sea, due to the extensive use of water The formation of mineral scale associated with the
injection for oil displacement and pressure production of hydrocarbon has been a concern in
maintenance, many reservoirs experience the oilfield operation. Depending on the nature of the
problem of scale deposition when injection water scale and the fluid composition, the deposition can
starts to breakthrough. In most cases the scaled-up take place within the reservoir which causes
wells are caused by the formation of sulphate scales formation damage or in the production faci1itie~
of Barium and Strontium. Due to their relative where blockage can cause severe operational
hardness and low solubility, there are limited problems. The two main types of scale which are
processes available for their removal and the commonly found in the oilfield are carbonate and
preventive measure such as the 'squeeze' inhibitor sulphate scales. Whilst the formation of carbonate
treatment has to be taken. It is therefore important to scale is associated with the pressure and pH changes
have a proper understanding of the kinetics of scale of the production fluid, the occurrence of sulphate
formation and its detrimental effect on formation scale is mainly due to the mixing of incompatible
damage under both inhibited and uninhibited brines, ie. formation water and injection water. In the
environment. North Sea, the universal use of sea water injection as
the primary oil recovery mechanism and for pressure
In this paper, we will present results of BaS04 maintenance means that problems with sulphate scale
formation kinetics in both beaker tests and in highly deposition, mainly barium and strontium, are likely
reproducible sandpacks which simulates the flow in to be present at some stage during the production life
porous medium. The effect of scale depOSition on the of the field.
dynamics of formation damage will also be
discussed. In the studies of BaS04 crystal growth Apart from its likely occurrence, the relative hardness
kinetics and formation damage, we have included and low. solubility of the sulphate scale means that
both normal formation/injection brine mixture and the very few remedial treatments are available for its
addition of scale inhibitor chemical. There are removal. Processes like acidization which can
significant differences in the results of static (beaker successfully remove carbonate scale, ego CaC03,
tests) and dynamics (core flood) conditions, and cannot effectively apply in this case. The common
between the normal and inhibited brine mix. practice in the oilfield to overcome the problem of
sulphate scale formation is by the preventive method
References and fi~ures at end of paper of 'Squeeze' treatment in which scale inhibitor
chemical is used to retard the kinetic growth of scale

429
2 Kinetics of BaS04 Crystal Growth and Effect in Fonnation Damage SPE 23814

crystal. Indeed most of the work reported to date value was based on the scaling tendency model [5]
[1,2,3,4] on the kinetic studies of BaS04 crystal whose prediction gave the highest degree of
growth are related to the screening of effective supersaturation with maximum amount of
inhibitor chemicals. In this paper we shall present the precipitation of BaS04 scale. Separate inlets were
results on the kinetic growth of BaS04 scale under provided for the different brines and these were
the influence of added sand particles, seeded crystal positioned so as to ensure that the mixing only took
and inhibitor chemical. Furthermore, results of in- place within the specific section of the column. A
situ scaling in highly reproducible sandpacks have schematic diagram of the in-situ scaling experiment is
been included to provide information on likely shown in figure {I}. Two different sandpacks,
formation damage and comparison of crystal growth using the same batch of BDH silica sand as in the
under static and dynamic flow environment. kinetic studies, were prepared for the inhibited and
uninhibited scaling experiments. The inhibitor used
was the sodium salt of the same phosphonate type
which, unlike in the previous case, contained
EXPERIMENTAL approximately 25% activity. The inhibitor chemical
was added to the sea water stream and gave an an
overall concentration of 50 ppm on mixing. The
In the crystal growth kinetic experiments, a combined injection rate of the two different brines
supersaturated solution was first prepared by mixing was 30 cm3/hr. This represented a residence time of
equimolar amount of BaCl2 and H2S04 solutions. approximately 55 minutes within the mixing section.
By controlling the rate of mixing and temperature Details of the sandpack assemblies and the injection
variation, ie. +/- 0.5 0 C, solutions of relative sequence are listed in table {2}. The scaling process
supersaturation of 5 could be preserved for a few was monitored by analysing the effluent
hours. The growth experiments were then initiated concentration of the scaling cations using Atomic
by the rapid addition of either sand particles, seed Absorption Spectroscopy (AAS) and, where
crystals of BaS04, or both. The process of growth appropriate, the concentration of inhibitor chemical.
was monitored by measuring the solution At the end of each experiment, the sandpack was
conductivity with time which was then converted to carefully dismantled and the spatial variation of the
the rate of change in scaling ion concentrations. The scale deposit was analysed using Scanning Electronic
seed crystals were previously prepared by mixing Microscopy (SEM). In order to provide direct
O.OIM of BaCl2 and H2S04 solutions over a period comparison of the scaling mechanism, static jar tests
of 24 hours before being ftltered and washed. The using the identical sea water and formation brine
final suspension contained approximately 39 mg of mixture, with or without inhibitor, had been carried
seed crystals per 1 cm3 of suspension fluid. The out.
sand particles used were clean, acid washed silica
sand supplied by BDH, of GPR grade which had a
mean particle size of 273 Ilm. The background
RESUL TS AND DISCUSSION
conductivity due to the possible surface charge of the
sand particles had been included in the calculations.
The value was obtained by measuring a standard 1. Crystal Growth Kinetics
solution containing 10 g of sand and 100 cm3 of
distilled water after it had been purged of dissolved The rate of crystal growth was monitored by
carbon dioxide by bubbling through with nitrogen measuring the solution conductivity. This was then
gas. In the case of inhibited crystal growth converted to the concentration of the scaling ions
experiments, the procedures were similar except that remaining in solution using the following equations
10 ppm of inhibitor was added to the original BaCl2 (6] :
and H2S04 solutions. The inhibitor used was a
commercially available phosphonate type called
Diethlenetriaminepenta (methylene phosphonic acid)
Il = Ilo - Kce n (1)
which contained 50% activity. k =:; CIl /1000 (2)

In the dynamic experiments, the in-situ scaling

process was investigated using reconstituted sea
where A. = equivalent conductivity
water and formation brine in the sandpack assembly. in ohm- 1m2equiv- 1
The compositions of the two different brines are 1..0 = equivalent conductivity at infinite
given in table 1. A ratio of 60% of formation water to dilution in ohm- 1m2equiv-1
40% of sea water was used for the experiments. This C = equivalent concentration in unit of

430
SPE 23814 R.M.S. Wat, K.S. Sorbie, A.C. Todd, P. Chen, P. Jiang 3

equivalents per litre, sites on the crystal surface by the inhibitor molecules
1O-3/C in m3 equiv- 1 [1].
k = conductivity in ohm-1m- 1
Kc = constant obtained from equation (1) For those experiments in which the crystal growth
process was initiated by heterogeneous nucleation,
with C and A. taken at 5x 10-4 mol/l ie. the addition of sand particles (expts. 30 & 53) or
a mixture of sand and seed crystals (expts. 55 & 58),
The accuracy of the above equations was confirmed the rate of growth can no longer be adequately
by constructing a calibration curve using different expressed by the equation (3). The results from
supersaturated solutions of H2S04 and BaCh. The these experiments, however, have been included and
conductivity of the calibrated samples was measured are shown in figures [5] and [6] in which
and compared with the calculated values given by the comparison between the inhibited and uninhibited
above equations and good agreement was obtained as growth can be made. In the case where growth was
shown in figure {2}. initiated by sand particles only, there was no
appreciable differences in the rate of growth. On the
For both inhibited and uninhibited growth, the other hand, significant rate reduction was observed
experimental details are summarised in table 3. In in the presence of inhibitor when the growth process
general, it is possible to prepare solutions with was initiated by both sand and seed crystals. This
relative supersaturation of 8 before any premature further indicates that the inhibition mechanism was
precipitation starts due to instability[7]. In all the mainly due to the blocking of active sites on the
experiments discussed here, solutions of relative crystal surface.
supersaturation of 5 were used and these were based
on the barium sulphate solubility taken as 1.009 x
10-5 mollL[8]. In the case where the growth was 2. In-situ Scaling Studies
initiated by the addition of seed crystals, the changes
in solution conductivity over time are shown in The pressure drop over the mixing zone of the
figure 3. As expected, the rate of crystal growth is sandpack was monitored throughout the experiment.
much retarded when inhibitor is present. No appreciable change was observed as shown in
Furthermore, results from both experiments exhibit table [2]. This was mainly due to the original high
the initial surge over the first 80 minutes which column permeability and the total scale deposit was
characterize the growth reaction from a metastable relatively insignificant when compared with the pore
supersaturated solution and is generally attributed to volume of the mixing section.
the surface nucleation process[9]. It is then followed
by a more gradual rate of growth which can be Apart from monitoring the pressure drop, effluent
represented by a second order rate equation samples were collected during the experiments. They
[4,7,9,11] in the form of: were analysed for the concentration of the two main
scaling cations, ie. barium and strontium, whose
dm effluent profiles are shown in figures 7 and 8
rate of crystal growth = -- ' = ks(m - m a )2 (3) respectively. In these figures, the effluent profiles of
dt
both the inhibited and uninhibited experiments are
shown. Also included in these figures are the [Ba++]
where m = molar concentration and [Sy++] concentrations measured during the static
IIIo = equilibrium (solubility) value jar tests. In the case of uninhibited scaling,
ks = crystal growth rate constant, a experiment ES-1, the [Ba++] ion concentration at the
function of the number of active outlet of the sandpack remains constant at
growth sites on the added crystals approximately 3.5 ppm throughout the experiment.-
This is consistently lower than that of the similar
In figure 4, the conductivity profiles are converted to beaker test even though the brines used and their
concentration and expressed as the integrated form of mixing ratio were identical. The increase in scaling
equation (3). From figure 4 the values of ks of tendency of BaS04 within the sandpack was likely
experiment 54 and 57 are found to be 0.0IxlO-2 and due to the abundant heterogeneous growth sites
1.26xlO-2 respectively. This indicates that the rate of which were available.
growth in an inhibited environment is about two
orders of magnitude slower than when no scale For the experiment ES-2 in which 50 ppm of
inhibitor is used. The retardation of growth is inhibitor was present, the results show quite a
generally attributed to the blocking of active growth different trend. In this case the [Ba++] ion
concentration in the sandpack effluent, after it

431
4 Kinetics of BaS04 Crystal Growth and Effect in Formation Damage SPE 23814

initially drops to a minimum of 14.5 ppm, rises and including careful examination of the SEM picture of
remains steady at a much higher level than the beaker the outlet filter and pressure monitoring, to suggest
test results. This indicates that the rate of BaS04 that homogeneous nucleation and subsequent particle
precipitation within the sandpack is somewhat slower blockage is the main contribution to the permeability
than that in static condition when inhibitor chemical impairment. The formation damage due to scale
is present. The initial surge of the precipitation rate, deposit is likely caused by the continuous growth of
however, as indicated by the drop in [Ba++] scale crystals which have remained stationary at the
concentration, is due to the non-equilibrium active growth sites. Finally, the differences in crystal
adsorptjon of the inhibitor chemical. When the sea morphology can be seen by comparing the pictures in
water containing scale inhibitor was first injected into figure 10. The scale crystal nearest to the point of
the sandpack, the advancing front of the inhibitor mixing (figure lOa) exhibits a very 'healthy' state of
chemical was retarded due to adsorption. Meanwhile, growth whilst for those which are further away
both the sea water and formation water, with (figures lOb and 1Oc) both the size and growth have
depleted inhibitor concentration, propagated down been somehow impaired. This is because near the
stream and resulted in extensive scale precipitation. point of mixing, the brines have the richest
The rate of precipitation gradually decreased as more composition of the scaling ions. The precipitation
free inhibitor molecules were available after the process is fast and is continuously supplied with
adsorption sites had been saturated. The delay of fresh brines. Further downstream, however, the
inhibitor breakthrough due to adsorption and the composition of the brines has been significantly
corresponding rise in BaS04 precipitation, as modified with the continuous depletion of the scaling
indicated by the drop in [Ba++] concentration, can be ions. This means that under steady state condition as
clearly seen in their breakthrough profiles which are during the sandpack experiments, there is a transition
shown in figure 9. Furthermore, apart from the initial zone of varying supersaturation along the sandpack
nonequilibrium process, the inhibitor effluent column and this gives rise to the different scaling
concentration remains constant at approximately 40 tendency.
ppm (80% injected value) throughout the experiment.
The loss of material during this steady state is likely
due to a combination of adsorption and bonding
mechanisms of inhibitor molecules with the newly
formed crystal surface. CONCLUSIONS

The differences of SrS04 scale precipitation under

dynamic and static conditions are less pronounced as In the current studies, we have presented results
shown in figure 8. Indeed the final [Sr++] which are used to quantify the kinetics of BaS04
concentrations in the sandpack effluent and beaker crystal growth under a controlled environment. The
test samples agree well with each other for both the work also extends to cover the realistic system in
inhibited (expt. ES-2) and uninhibited (expt. ES-l) which in-situ scaling takes place under dynamic flow
cases. The lack of sensitivity of strontium scale conditions. Based on the results obtained from these
precipitation is mainly due to the fact that the studies, the following conclusions can be made.
supersaturation of SrS04 is about one order of
magnitude smaller than that of the BaS04 [5] in the
(1) The rate of growth of BaS04 scale, after the
brine mixtures used for the experiments.
initial surge, can be represented by a second
Other important information on the in-situ scaling order rate equation. However, this does not
mechanism is the morphology of the scale deposit. In extend to the case when crystal growth is
figure 10, the SEM pictures of the uninhibited scale initiated by heterogeneous nucleation.
deposit (ES-l) at three different locations of the
sandpack are shown. They are taken at 2.5 cm, 12.5 (2) When scale inhibitor is present, the rate of
cm and 38.5 cm respectively from the point of growth is significantly retarded and the rate
mixing. A number of interesting observations can be constant, ks, is approximately two orders of
seen from these and many other similar SEM magnitude smaller than the uninhibited case.
pictures. Firstly and as expected, the amount of scale The differences in rate will vary according to
deposit is most abundant near the point of mixing the fluid system and the experimental
and generally decreases with distance. Secondly, conditions.
perhaps more importantly, the scaling mechanism is
very 'localised' with mainly heterogeneous growth (3) For the in-situ scaling experiments, the
on sand grain surface. There is no strong evidence, crystal growth process appears to be
localised. The permeability decline is likely

432
SPE 23814 R.M.S. Wat, K.S. Sorbie, A.C. Todd, P. Chen, P. Jiang 5

caused by the continuous growth of crystals 4. Liu, S. T. and Nancollas, G. H., " The
and not by particle transport and flow crystal Growth and Dissolution of Barium
blockage. Sulphate in the Presence of Additives", The
Journal of Colloid and Interface Science,
(4) The scale precipitation process in porous Vol 52, No.3, 9,1975. 583.
medium is likely to be dominated by
'heterogeneous nucleation with the [Ba++] ion 5. Yuan, M.D. and Todd, A.C., "Prediction of
concentration in the sandpack effluents Sulphate Scaling Tendency in Oilfield
~onsistently less than that of similar beaker Operations", paper SPE18484 presented at
tests. However, when the process is inhibited the SPE International Symposium on Oilfield
by the addition of scale inhibitor, the extent Chemistry, Houston, Texas, 8-10 Feb 1989.
of inhibitor adsorption on the sand surface
will significantly modify. the process. The 6. Moore, W. J.," Physical Chemistry", Lowe
effluent [Ba++] concentration in this case is & Brydone Ltd, London 1968.
higher than the beaker test which indicates
less precipitation in-situ. 7. Nancollas, G. H. "Crystallisation of Barium
Sulphate in Aqueous solution" Trans of
(5) The scale morphology varies considerably Faraday 59,1963. 737.
from the point of mixing. Under steady state,
the scaling process within the porous medium 8. VanDer Leeden, M. C. and Van Rosmalen,
is contributed by a range of supersaturated G. M., " The Influence of Various
mixtures. The point where the incompatible Phosphonates on the Growth Rate of Barium
brines first come into contact with each other Sulphate Crystals in Suspension", Estudios
has the highest supersaturation and with Geol, 38, 1982, p279-287.
maximum amount of deposit. This is
followed by a transition zone of rapidly 9. Leung, W, H. and Nancollas, G. H.,"A
depleted scaling ions. kinetic Study of the Seeded Growth of
Barium Sulphate in the Presence of
Additives", J. Inorg. Nucl. Chem. Vol 40,
p1871.

ACKNOWLEDGEMENT 10. Liu, S, T. and Nancollas, G. H., " The

crystal Growth and Dissolution of Barium
We would like to thank the member organisations Sulphate in the Presence of Additives", The
funding the Heriot-Watt University Oilfield Scale Journal of Colloid and Interface Science,
Research Project which includes B.P., Chevron Vol 52, No.3, 9,1975. 583.
U.K. Ltd., Ciba-Geigy Additives, Marathon, Shell
UK Exploration and Production, Statoil, Texaco, 11. Rubin, A. J. , "Aqueous - Environmental
and M.T.D. Ltd/SERC (Grant no. GRIE63322). Chemistry of Metals", Ann Arbor Science
Publishing Inc, Michigan 1974.223-224.
REFERENCES

1. Nancollas, G. H., " Oilfield Scale - Physical

Chemical Studies of Its Formation and
Prevention", Chemicals in the Oil Industry
1985, 143-164.
2, Vetter,O.J., "An Evaluation of Scale
Inhibitors", Journal of Petroleum
Technology, p997, August 1972.
3. Boyle,M.J. and Mitchell, R.W., "Scale
Inhibition Problems Associated With North
Sea Oil Production ", SPE 8164, presented
at the Offshore Europe 79 Conference,
Aberdeen, Scotland, 3-7 September.

433
 SPE 2381 4

Table 2 Summary of Sand pack Experiments

Sand pack Experiment ES-l ES-2

T->O ~~ }>O It>O
[Column Length (cm)
Table 1: Composition of Formation and Injection Sea Waters 69.5 41.8 69 40
fPore Volume
(ml) 45.47 26.32 48 29
Ions (mglL) Jo'ormatlon water inJection sea I Sand weIght
water (2) 205.50 199.76
Na :lY,40C 1 J,H9l [Dead Volume
..:a 2,8OC 428 (ml) 1.0 1.5
Mg :lot 1,368 I PorosIty 0.37 0.3~
K 37C 46l I Permeability (d) A->B B-> A->B B->
Sr 575 Initial 24.5 ~.nr 8.97 39.58
J:la 25 Final -25.5 29.31 .64 41.52
1~,/{)t [Floods Sequence: 1) tracer, U (1->0) 1) tracer,
~
52,64 (1->0
::iU4 2,96l 2) tracer, U (B->O) 2) tracer, (B->O)
3)FW& (1->0) 3)FW& (1->0)
sw (B->O) SW& Inhibitor (B->O)
4) tracer, U (B->O)
[Flowrate
(ml/hr) 30 60 30
ResIdence T,me
(min.) 52.9 26.5 58
Pore Vol. Injected
26.65 79.95 74

Table 3 Summary of Crystal Growth Kinetics Experiments

Expt. super- Initial 11 Final 11 Seed Sand pH ks(10- 2 )

saturation cone. cone. Crystal (g)
oom oom (ml)

30 5 5.00 4.0

57 5 1.00 4.0 1.26

58 5 1.00 5.00 4.0

53 5 10.00 9.76 5.00 3.73

54 5 10.00 9.57 1.00 3.76 0.01

55 5 10.00 9.48 1.00 5.00 3.75

Ip, I~
manometer
I Pd

-.
25cm 15 cm
FW 18ml/hr.

mpump
I A

~
B ,
, ,,
,,
D 0

~~~~
sw
12mIlhr.
,,
,
,,

B section
sample
collector

inlet & Pb

'"'" --+_'"'01
inlet

Fig. 1-Apparatus layout for In-situ scaling experiments.

434
 SEE 2381 4

80 7

70
6
~

;::;
60 ~

a
y
a 5
~ 50 I/)
::l
Q
&.
...,:; 40 2 4
]y .S
I:

=
"0
I:
~
30 .
-;
3
U
0 measured values
-='"y
20 I:

10
-- calculated
U
~

2 • uninhibited
0
10 ppm inhibitor (Expt.54)
(Expt. 57)
O~~--r-~~r-~~--~-r--~-r--r-,-~~
1+-~--.-~--~~~~--~_r--~~--~~
0.00e+0 1.00e-5 2.00e-5 3.00e-5 4.00e-5 5.ooe-5 6.00e-5 7.00e-5 o 50 100 150 200 250 300
Concentration, mol/l
Fig. 2-Predlcted and measured conductivity of supersaturated solutions of H2 S0 4 and BaCI 2 •
Time (min)
Fig. 3-Conductlvlty profiles of Inhibited and uninhibited crystal growth experiments Initiated
by adding seeded crystals.

40oooT-------------------------------------~ 6.5 ,-----------------------------------------..,

• IO ppm inhibitor (Expt. 54)

o uninhibited (Expt. 57) • IO ppm inhibitor (Expt.53)
6.0 o uninhibited (Expt.30)
30000 s
~
a
5.5
20000
I:
~

..
.~ 5.0

-=y'"
I:
~
U 4.5

150 200 250 300

4.0 +--~__,-~----r-~-_._-~-..._---,.-~-~---l
Time (min) o 50 100 150 200 250 300

Fig. 4-Rate of change of scaling Ion concentrations In Experiments 54 and 57. Time ( min)
Fig. 5-Concentratlon profiles of scaling Ions remain In solution after the addition of sand particles.
 7 150
o Ba (15Oppm on mixing, no 12)

6 • 10 ppm inhibitor (Expt.55) 125 • Ba (15Oppm on mixing, with 50ppm 12)

--Ei
~

'0
o uninhibited (Expt.58) Ba (beaker test, no 12)
Ba (beaker test with 50ppm 12)
5 100
III
.;.
~

6 4
75 •
i
..•..............
c:
.~ Q,
-; ~
.......c:
..c:
.
3

• ••••••
<II 50
=:I
u
e J8ppm
U 2 25 /
---------------------------------------11-----------
·----·-0--0---0--···-0-····-0-·--0-:-·0···-····0·-- , 8.6ppm
1 0
0 50 100 150 200 250 300 10 20 30 40 50 60 70
0

PV (mixing zone) Injected

Time (min)
Fig. 7-Eflluent concentration prolilea of (Be + + I for inhibited and uninhibited In-sltu scaling exparimente together
Fig. 6-Concentratlon profiles of scaling Ions remain In solution after the addition of seed crystals and sand particles. with the reaulta from the parallel beaker teate.

3ro~--------------------------------------, c: 1.0
Sr (34Oppm on mixing, no 12) .~
~

,
-;
Sr (340ppm on mixing, with 50ppm 12) ...
340 Sr (beaker test, no 12) =..c:
u
0.8 •• ooCJXJr::p:jJ 0
0 0

Sr (beaker test,with 50ppm 12)

e
u ~ cPcP
.... 0.6
..=1:1 ~
• 00
320
IE.. 0
~.~..
____
"» 6 Iii.
•
t j l•••••
___________________________________
. . . . ••• 41.. ________ _ I 300ppm .
"CI
.~
0.4
• 0
0

';

••
0.2
300 ~ II ,'.I"IUIIMh~ III
...Eie m l-
• c:
0.0
O~
280 0 2 3 4 5 6 7

pore volume (mixing zone) injected

2W+-~--.-~~r-~_.--~_r--~_.~--._~__4 Fig. 9-Braakthrough profiles of Inhlbltor and [Be + + I for the Experiment E8-2.
o 10 20 30 40 50 60 70

PV (mixing zone) Injected

Fig. B-Effiuent concentration proliles of (Sr + + I for Inlllblted and unlnhlblteclln.aftu scaling exparirnanta together
with the reaulte !rem the parallel beaker testa.
 SPE 2381 4

A B

Fig. 10-SEM piclures 01 scale cryslal lormed al (a) 2.5 cm, (b) 12.5 cm, and (c) 38.5 cm Irom poinl 01 mixing In
Experiment ES-l.