, ● ✎✌

Mechanism of the Dissolution of Salt.. in the Formation

of Underground Salt Cavities
R, W, DURIE* ,.,<
JUNIOR MEMBER AIME
F, W, JESSEN
MEMBER AIME I THE U. OF TEXAS
AUSTIN, TEX.

develop maximum storage volumes within specified

ABSTRACT
regions.
The development in the laboratory of small salt h is the purpose of this paper to discuss some
cavities illustrated that the rate at which salt aspects of the mechanism of the dissolution of
dissolved was affected by the rate at which fresh salt in the formation of cavities of controlled
water was injected into the cavity. Experiments geometry. Although no attempt is made to resolve
carried out with small sait cores indicated, however, the” problems that may be associated with the
that the rate of salt removal with forced convection storage of tnethane in the liquid srate in salt
was not significantly dif~ererzt from that with natural formations, the fact that salt cavities can be
convection, provided the flow was in the laminar formed at reasonable cost would appear to justify
range. The rate of salt removal was apparently investigation in this direction.
controlled by the water salinity, which was indirectly Further, underground cavities have the advantage
determined by the nzte at which fresh water was of allowing storage at higher pressures (and hence
inject ed. With t urbrdent [low at the salt srojace, higher temperatures) than surface installations.
the rate of salt zemoval could be increased several- Although the thermal conductivity of salt is some-
fold with the same salinity $n the cavity. what higher rhan frozen earths, 0.017 caI/(see)
A quantitative evaluation of the rate of salt (cm)(Y) for salt at 32F vs 0.005 caI/(see) (cm)
removal under natural convectiofl was attempted (%2) for ice at this temperature, this disadvantage
by using techniques developed for the analogous is offset by th~ higher storage temperature in
beat convection system, The values of ~be rate oi underground cavities.
salt removal obtained experimentally were consider- For example, at atmospheric pressure, the
ably greater than the calculated values due to temperature of liquid methane would be -258F. In
sur/ace irregulrzrities. To determine the e//ect of a storage cavity at 2,000 ft, the pressure would
these irregrdarr”ties, the rate of salt removal was be approximately 1,000 psia, whi~h would allow
found fcr a smooth surface by experimentally a storage temperature of -901?. The formation
determining the rate of salt removal as an irregular temperature would be roughly 110I? at this depth,
surface developed and then extrapolating this hence the total temperature differential would be
trend back to initial time. Excellent agreement 200F. Under surface storage conditions the
between the experimental and calculated values corresponding temperature difference would be
then were obtained, some 333 F (75 F + 258 F). The results of this dif-
ference can be compared more readily by substituting
JNTRODUCTION the appropriate values of thermal conductivity,
temperature difference, heat capacity and density
Underground cavities leached from massive
into- the solution of the diffusion equatiqn for the
salt formations have been used extensively for unsteady linear conduction of heat. ‘<
many yeats for storing LPG products. ‘The develop-
For the underground salt storage, the heat fIux
ment techniques for storing methane gases in the is 15/~caI/(see) (sq cm), where t is the time in
liquid state at a low temperature leads to spectdacior, seconds from’ the initiation of the cooling. For the
that underground salt cavities of controlled geomet, ~, surface system, and using the thermal properti~s of
mav. be suite
. im~ortant
. in this field, In any case, ice at the average temperature, the corresponding
the increasing number of salt cavities used .f~r ~
figure is 21/@ cal/(see) (sq cm). Hence, the h$at
storage indicates an irnpendhg riced for a betr,’ v ~ losses” through” ice undkr surface “conditions would
knowledge of the washing process to efficiently’ ‘ be some .30 per cent higher than the underground
Orieinal dwmuscript received in S?clety of Petroleum skit ..titorage, This ratio would vary considerably,
Ensineers office July 19, 1963. Revised manuscript received of course, for other conditions, but it indicates
April 6. 1964. Peper preeented et SPE Fell Meeting, held In
New Orleans, Oat. 6-9, {963. that an advantage over surface facilities may be
@Tfowwith Shell Canede Ltd. in Edmonton, Alta. obtained by storage in underground salt cavitiesi
.,. 1ss

----- ... . .. . ... . . . . . ~. ---- . . . . . . ... . . ,. ..:

 .“

Spherical salt storage cavities would minimize INDUCED FLOW EXPERIMENTS

heat transfer to a stored liquid because a sphete
presents the least surface area for a specified
storage volume. Controlled geometry is particularly
important, therefore, when liquid methane storage
is considered. k is possible to leach model salt
cavities in the laboratory with a spherical shape
using a blanket of gas or hydrocarbon liquid on
top of the water inthe cavity to control the washing
process in the upper regions, and the adjustment
of the injection rate and the position of the tubing
strings to control the development of the low-cc
regions.l, s Such a cavity is shown in Fig. 1.
However, extension of these techniques to field
prototypes requiresa good understanding of the
washing mechanism.
The rate at which fresh water is injected into a
cavity affects the shape of the cavity and the
rate at whichsalt isremoved. This can be attributed
‘to the saIiniry profile that exists, and not directly
to the flow velocity within the cavity. If a cavity
were filled with fresh water and the rate of injection
were very small, then the warer would not carry
the salt out through the tubing as rapidly as it
dissolved. The salinity in the cavity would begin
to increase and hence the rare at which salt
dissolved would decrease until some equilibrium
was established. There probably would be a less
saline water toward. the top, so the rate of washing
would be greater near the top than in the Iowe;
porrion of the cavity.
If the rate of fresh water injection were increased,
a new equilibrium saturation “would result, which
in turn “would cause the salt to be dissolved more
rapidly. However, the upper limit for the rate of
dissolving the salt would be essentially the rate
at which salt dissolves under the influence of
convection in fresh water, since significant linear
velocities would not be attained? If the conditions
of flow in the model or field prototype were sufficient
“to cause turbulence, the rate of dissolution would
then deviate very significantly from that for natural
convection.
LRefarenCe~ given at end of paper.
Preliminary experimental work was carried out
to investigate the itrfluence of flow rate on the
rate of dissolution of small salt cores. Salt cores
2,5 cm in diameter and about 6 cm long were cut
from raw salt blocks from the Grand Saline, ‘Tex.,
mine of Morton Salt Co., using a diamond core bit.
The ends of each core were sealed to glass tubing
of the same diameter. For each experimet.tal run
the rod formed by the glass, together with the core,
waa mounted within a glass cylinder of 3.75 cm ID.
This was mounted in: a container for the flow test
so water could be dis@aced downward through the
annular space at very slow flow rate?. The volume
of the salt was measured by volumetric displacement
before and after each run, The average dimensions
of each core were found and the rate of salt removal
was calculated and expressed in units of cc/(sq cm)
(see).
Results obtained from several salt cores using
both fresh water and a saline water with a sal~
concentration of 1.8 moles/IiCer are illustrated in
Fig, 2. Considerable. scattering of the data points
appeared to be due to the development of an irregular
surface on the core. Despite the scattering, these
results indicated that the induced flow, at least
at the lower rates, ,does not contribute significantly
to the rate of salt removal. Core samples that were
dissolved with no induced flow, i.e., entirely by
natural convection, exhibited rates of dissolution
that could not be distinguished from those cores
dissolved with low induced-flow rares. It was
apparent that the significant factor affecting the
rate of dissolution of salt by forced convection
/$’
@&”
,/”
0
0
$ALINE WATER _——c
.—— ———
0
0
0

LEGEND
X FRESH WATER
O SALINE WATER
—.

I I i I i I I
1 234567
AVERAGE LINEAR FLOW VELOCITY cm/sec
FIG. 2 — DISSOLUTION RATE OF GRAND SALINE
. .
FIG. 1 —LABORATORY fjAVKTY, SALT CORES.

.;1s4
,.,
... ... . . . . . ... .... ..... . .. . .-... . . .. _ .. SiOCIETY
. ... -----
OF -.:-
PETROLEUM-- . ENGINEERS
-.: - -JOURNAL
---–-
—.
 . ... I

with Iaminar flow was the salinity of the water TASLE 1 - DISSOLUTION DATA FOR CONTROLLED
and not the flow rate. FLOW RATES*
The surface irregularities that developed in the Inlti Rate of Salt Re oval
Core Llnoor
salt dissolution experiments” appeared to vary Lenmh I!?cc/(sq cm) (see)
Flow Velocity
between samples. It was decided that the random (c;) Experimental Ad[usted (cm/see)
——
scattering of data points could best be reduced by 3,80 0175 ,178 o
eliminating the effect of rhe surface irregularities. 3066 ,190 ,190 0.1
This was done by an experiment similar to that 3,50 h178 .177 O*4
previously described, but the v@rme of each salt 3,58 ,370 ,370 9.4
core was measured at several intervals during ●Wlnnfleld Salt Cores -- Fresh Woter
the dissolution cycle instead of only at thebeginning
and end, This allowed investigation of the rate of
0.OO6 cm/see, which is well below the velocities
dissolution vs the dissolution time as illustrated that appear to disturb the natural convection
by an example, Fig. 3. The rate of salt removal boundary layer.
increased as the surface irregularities developed,
and it was possible to extrapolate this trend back MATHEMATICAL ANALYSIS
to the initial time to obtain the rate of dissolution
for a smooth surface. Natural convection effects are of prime importance
Table 1 records rh; results from four cores in the analyais of controlled dissolution of salt.
investigated in this manner. Initial dissolution .Therefore,, it was considered important chat a
rates were nearly independent of the flow rate method be developed to evaluate convection flow
for flow in the Iaminar range. The higher flow rate with water exposed to a vertical salt surface and
of 9.4 cm/sec corresponded to a Reynolds number hence determine quantitatively the rate of salt
of 5,000, hence turbulent flow could be expected removal into water of a known salinity,
to affect the rate of dissolution for this sample. Water in contact with a salt surface will increase
To investigate turbulent flow, water was flowed in density as salt is dissolved and will move down
at high races through 1.6 cm diameter holes in the surface by natural convection, The velocity
salt blocks, and the rate of salt removal was found of this flowing boundary layer is, a function of
from weight measurements. With initial flow the density difference. between the boundary layer
velocities of 150 and 340 cm/sec the ;ate of salt and the main body of water, as well as the viscous
removal was 10 to 20 times that observed us”ing shear forces that tend .to retard the flow, It has
laminar fIow rates and small cores., been shown by Eckert2 that the temperature profile
“ The results of these experiments lead to a better is parabolic in the analogous heat convection. The
understanding of the effect OF flow velocity when concentration is therefore assumed to be parabolic
salt cavities of specific geometry are being leached. (see Fig, 4).
For example, in the development of spherical The coordinate axis used in” this analysis has
laboratory models, the flow volumes are generally the x-axis parallel CO the salt surface, and the
in the order of 100 CO 200 cc/rein. If a volume of positive y-axis represents the horizontal direction
200 cc/rein flows through a circular section with a CONCENTRATION PROFILE
diameter of 25 cm, then the average velocity is A
4

I===y
SALT
4
“x
WATER

VELOCITY PROFILE

SALT
4 WATER

II
o
TOTAL

FIG. 3 — FREE
4
IMMERSION’

CONVECTION
8
TIME

DISSOLUTION
12 16 20
MINUTES - CORE 34

OF
-
“4
F’
.lX
FIG. 4 — CONCENTRATION
“
AND VELOCITY PRO- I
WINNFKELD SALT CORE IN FRESH WATER.

JUNG 1964 ‘ ., 18s

— . .-
 . .

away from the tmlt. If the line y = O represents the where p/3C is the den sit y difference.
salt surface, then the coordinate axis must shift to Neglecting acceleration, as discussed before,
the left as the salt dissolves, which is equivalent the momentum change equals force due to gravity
to a component of flow in the y direction. However, minus viscous shear force:
the flow in the x direction is in the order of 104
times the flow in the y direction, and hence the y
component can be neglected. Therefore, the flow
can be aasumed to be parallel to the salt surface.
The concentration and velocity profiles are
described by the following equations:

or:
c=c~(l-.+p . . . . * . . (1)

u=@f/_f)~: . . . . . . (2)
. . . . . . .. 4.,. .**QQO ● . (3)

dere C= is the difference between ch e salt con- From Eqs. 1 and 2 it can be shown that:
centration in the water at immediate contact with
the salt surface and the concentration of the main 8
body of water. U1 is an unknown constant to be / y=o
U2dy=~
determined and 8 is the width of the boundary
layer. It will be necessary to assume that the width
of the region of flow (the velocity boun,dary layer) is 8
the same as the width of zone of concentration f y=o
C dy=~
change (concentration boundary layer). Although
this will not be the case in a viscous Iiquid, this
wiIl not affect the calculated dissolution rate,
provided Eqs. I and 2 adequately describe the
concentration and velocity profiles close to the
salt surface.
Then by substitution into Eq. 3:
tiOMENTUM EQUATION

The forces acting on the boundriry layer described

above are equated to obtain a momentum equation. .. . . . . . . . . . . . . . . . . . . . (4)
The mass flow per unit time through a segment of
width Ay and unit thickness is equal to uAy, and SALT TRANSFER
the momentum is the product of the mass and velocity
. U2Ay.. Then the equation that describes the The rate of salt transfer by diffusion in water is
determined by the concentration gradient and
total momentum is
diffusivity constant. For a segment’ of imit width
and height Ax:
P/;2dy
y=o
The momentum change with progression in the x
direction is described ,by:

dX (~’..
d 8U2dy}M
The momentum change is opposed
force T5 at the salt surface:
, The rate of salt dissolution will, of course, be
controlled by the rate of salt transfer at the
water-salt interface. If the boundary layer flow is
by a shear laniinar, mixing can be neglected and the rate at
which salt is dissoIved will be:

—. The force of gravity acting on an element of the

botitidary “Iayer to””cause” “flow is g x mass, and
()
dC’ ”’”.
where — is the concentration gradient at the
for t~e total boundary layer is: dy s
water-sale contact.
In an element of the boundary layer “of unit ‘
. thickne SS, the increase in -salt content is KC,%ZAy.
Then in a segment of height Ax the sah content ;

.
.. ...----,.
.-. . . .. . . ... . . ... .
--- . *.. . .- -.

is:

. . . . . . . . . . . . . . . . . . . . (9)
The rate at which salt is carried into this elemenr AQ-2KOfSAx
depends on the flow velocity, as follows:
At
8 , . . . . . . . . . .“ . . (10)
A(?l= UCdy At,
KJ=O

‘“ The rate at which salt is carried away is described The factors used to simplify the above equations
by a similar expression for AQ2. The net effect is: are listed in Table 2.
From Figs. 5 and 6 it will be noted that v/D is
AQ=AQ2– AQI the order of 1,000. Hence no appreciable error would
occur by substituting
hence

With further substitution of the above consrants,

Eqs. 8’ to 10 simplify as follows:

Equating Eqs, 5 and 6 yields

Um = 4.6(*)? Crnlsec . . . . . (11)

esta descreve
It can be ‘shown from Eqs. i and 2 that

b’
CU dy = ‘3+ u, Cs $ ,
f Y=o
—
-.
—
— . (d) . –pa p 3/

dy S B: (~)x =0.0732W~~ ~ 4$ gm/(sqcm)(sec)

,., ... .,, ,,. . . . . . . ,. (13)

Hence The above equations describe the dissolution

process as a function of water salinity at any
point on a vertical salt surface.
&#c~u,8 )=2&7+ To find the amount of saIt removed from a surface,
it is necessary to integrate Eq. 3 with respect to
... ,,. . . . . . . . . . . . . . . (7) x over the total height of the salt surface. This
yields
PERFORMANCE EQUATIONS

the analogous heat convection

In U1 and 8 are
exponential functions of x. That is,
gm/(sq cm)(sec), , . . . . “. . . ; . . . (14)
-, u~ = Clxm 8. = C2 Xn,
where H is the heig”hr of the exposed salt surface
U 1 and 8 as defined above can be substituted
into Eqs. 4 and .7. Since these equations must
both be vaIid at the same value of x, it is possibIe
TABLE 2 — FACTORS USED TO SIMPLIFY EQS. 8,9 AND 10
. . -to, solve for m, n, Cl and C2 as shown by- Eckert. —- .- . . ..
Using this approach it can be shown that Factor Volue
980 cm/sec2
: From Fig, 5
Urn = 0.776 u (0.952 +h(p) ‘X % K 0.0585gm liter/(moie) (cc)
From Fig. 6
. . ...’.. . . . . . . . . . . . . . (8) ;..
0,0367 gm liter/(molo) (cc)
 —

two Grand Saline cores dissolved by

natural
convection in fresh water were both x 10-3
0,160

dq .+ cm compared to theoretical values of 0.165 x IOA

and 0.167 x 10-3 cc/(sq cm)(sec). A third core
dt 3 dt
dissolved in saline water had theoretical and
experimental values of 0.0255 x 10-3 and 0.0265
—
— The factors D and v in the above equations are x 10-3 cc/(sq” cm)(sec) respectively.
used as constants, although they vary with
The dissolution of the small salt cores showed
concentration. However, the variation in D is not excellent agreement between the theoretical and
large and v is raised to a 1/4 power. Therefore, experimental values of salt removal rates from a
average values of D and v can be used. The smooth surface. However, these samples were too
maximum linear velocity, the thickness of the small to allow large surface irregularitlea to
boundaw [ayer, and the rate of saIt removal then develop, Therefore, several rectangular samples
are des;ribed as functions of C~ and %. with dimensions about 6 x 6 x 18 cm were prepared.
Exposed vertical surfaces were dissolved “ by
NArURAL CONVECTION EXPERIhfENTS natural convection ‘and the progression of the
dissolution was found by weighing the samples.
The above mathematical analysis indicated salt
Typical results are” illustrated by Fig. 7. Initial
dissolution rates were considerably less than
values measured with the salt cores, due to the rates of salt removal for three types of salt are
compared to a theoretical curve in Fig. 8. Good
development of the surface irregularities, However,
the experimental values for the initial smooth agreement is observed for the three types of salt
surface, found by the extrapolation method previousl y used. Experimental data together with values
described, showed excellent agreement with the calculated by Eq. 13 at the beginning and end of
calculated values, Initial rates of salt removal for the experiment are recorded on Table 3.
,/ The surface irregularities that developed during
‘~ I.8 the washing process appear to be caused by the
invasion of ‘water between salt crystals. This effect
j “s I was most pronounced in the sample shown in Fig.
/ <
E
“
1,4
9, a. Grand SaIine block exposed to fresh water
.& ~ I for 45 minures.
Fig. 10 shows a Grand Saline sample that was
immersed in water of salinity 4.02 to 4.55 moies/
liter for 679. minutes. The surface irregularities
are much less pronounced and the outer profile
HH+-HI1
g 0.4
II 11”{
, I
agrees very weIl with the theoretical profile.
The maximu’m rates of salt removal found
k .0.2 -
experimentally are compared graphically to the
n theoretical rates of Fig. 11. The maximum rate of
o salt removal is two to three times the theoretical
0 Lo 2L! 3!3 4,0 5.0 6,0
rate, as shown on the last coIumn of Table 2. The
CONCENTRATION MOLES PER LITER
ratio of the final average rate of salt removal to
FIG. 5 — DIFFUSIVITY CONSTANT OF AQUEOUS the theoretical rate was plotted graphically vs the
SODIUM CHLORIDE SOLUTION AT 18C.

.018

.016

.014 T

,012 /“
/ -
.010 — — - “

.008
“2

F-
.006
$- ,05 \
.004 - o Calculated
w
.-.,., ,002 .. . . . ._ _ 1-
u- .
a
0 I I I I
o “o 15 50 ?5 100 125
0 i.o 2,0-. 3.0 4.1) 5.0 T6TAL IMMERSION TIME MINUTES - SAMPLE 25
CONCENTRATION MOLES PER LITER
FIG. 7- FREE ‘ CONVECTION DISSOLUTION”’ OF
:’ kTGo 6. — KINEMATIC VISCOSITY OF AQUEOUS GF:AN.D SALINE SALT BLOCK, SALT CONCENTRATION
80DIUM CHLORIDE SOLUTION AT 20C, 2,5 TO 2.83 MOLES PER LITER.
.

.. . .
.---.
.
,.

TABLE 3 — CONVECTION DISSOLUTION DATA

Molar
●
Rate of Salt Remcwol [cc/(sq cm) (SOC) x IOS]
COncentratlOn Initial Final Average
Samp10 Length Total Immersion
Number Salt (cm)’ Time (minutes) Initlnl Final Experimental Calculated Experimental Calculated Rat io*
—— — —.
26 G. s. 17,3 679 4,02 4,55 ,028 .0236 ,026S *O130 2,04
25 G, S, 17.7 105 2,60 2.83 ,068 ,05S6 ,115 .0501 2.29
27 G. S. 14.7 75 1,69 1.90 ,106, .0855 .175 .0795 2,39
29 G. s. 14.4 35 0090 1.02 ,085 ,1040 ,460 ,0978 4*7O
23 G. s* 17.3 45 0 0 .150 ,1240 0550 . i 240 4.44
31 Hoc. 19.2 37a 3,72 3.92 ,021 ,0304 .057 ,0262 2,18
32 Hoc, 19*5 160 2,47 2*6S 0055 .0608 0118 ,0562 2,10
30 . HOC, 19.6 70 0 0 .115 ,1240 ,350 .1240 2,82
38 Wlnn. 18.9 503 3.71 3,93 00215 00305 ,057 ,0257 2,22
39 Winn. 19.3 305 2.33 2.41 .052 .0637 *149 .0620 2,40
43 WInn. 19,4 115 0,95 1,24 ,095 ,1000 ,250 .0919 2,”72
40 Wlnn. 19.6 120 0 0 ●115 ,1230 ,430 .1230 3,49

*Ratia: Final average rate of dlssalutldinltlal calculated rate.

—.
water salinity. A grouping was obtained that could CONCLIEWNS
be approximated by a satisfactory linear relation-
ship. The equation for this line is: 1. The experimental results substantiate the
theoretical values of the rate of dissolution of
A=/*70+ o.26cs . . . . ..(l5) salt from a smooth vertical surface by natural
convection. The adjustment factor gives a good
where A is the adjustment factor relating the final
approximation after the development of surface
average and theoretical rates of salt removal, The irregularities. This supplies a range for the
curve obtained by applying this adjustment factor
quantitative evaluation of salt removal from salt
is compared to the. measured values in Fig. 11, cavities. Overburden pressures, temperature effects
The , Grand Saline values that deviate from the and the absence of physical movement may cause
curve at low water salinity were neglected in
the degree of surface distortion to be less in
obtaining these relationships because of anomalous
underground cavities than in the Iaborator y samples.
removal of small salt fragments during dissolution.
2. Small, induced Iaminar flow rates past a
,.,
W

LEGEND
X GRAND SALINE SALT
O HOCKLEy SALT
A WINNFIELD SALT
fa

\ x’

0)

-J

*, “s
,.

,?
I I
I 2 3 4 “5
SALT SALINITY MOLES PER LITER

FIG, 8 — FREE CONVECTION DISSOLUTION OF SALT. FIG. 9 — FREE CONVECTION DEW.OJ-UTION

, OF
EFFECT OF WATER SALINITY ON INITIAL RATE OF GRAND SALINE SALT BLOCK IN FRESH WATER.
SALT REMOVAL.

:,~~~, 1964 .–-– :. :-- ‘-; . ‘ :- : ‘-. --.:

----- . . . .. . . ..!... .. .. .. . .. .. . .. . . . . . ..”. .
,.-
 LEGEND
X GRAND SALINE SALT
O HOCKLEY SALT
A WINNFIELD SALT

\
ADJUSTED THEORErlCAL RATE

\ x

\
\ xl
~
o I 2 3 4 5
SALT SALINITY MOLES PER LITER
FIG, 10 — comparison OF FINAL SALT FIG. II — FREE CONVECTION DISSOLUTION OF
TO PREDICTED PROFILE. SAMPLE 26. SALT, EFFECT OF WATER SALINITY ON MAXIMUM
AVERAGE RATE OF SALT REMOVAL,
salt surface. will not appreciably affect the rate of
dissolution of salt. The controlling mechanism is U = velocity, cm/sec
the convection flow process and the rate of dis- u~ = unknown function of velocity, cm/sec
solution of the salt is hence determined by the water Um= maximum velocity, cm/sec
salinity. The initial rate of dissolution at a point on V = viscosity, gin/cm sec (poise)
a smooth salt surface may be C1Osely approximated = kinematic viscosity, sq cm/sec (stokes)
by a theoretically developed equation.
~ = conversion from concentmtion difference to
3. Test s~ples became irregular as the washing
specific gravity difference, 1
progressed, and the rate of dissolution increased
due to the increase in the exposed surface area. moles/liter
The equation for a smooth salt surface can be 6 = width of the boundary layer, cm
adjusted to describe the final rate of dissolution p = density
of water, gin/cc
from the surface of small samples by multiplication P. = density of salt, gin/cc
.by an adjustment factor. T = shear force, gm cm/sec2 .,
4. If the flow rate by a surface is in the tutbulent
Ts = shear force at salt surface, gm cm/sec2
range, the boundary layer will be disturbed and the
rate of salt dissolution will be greater by an order ACKNOWLEDGMENT . .
of magnitude of several times that indicated by The authors wish to, acknowledge with thanks
the above equations. the support of the Atomic Energy Commission,
NOMENCLATURE whose sponsorship of a project carried out in the
c = difference between water salinity at a point Petroleum Engineering Dept. of The U. of Texas
in the boundary and the salinity of the made possible this research.
main body of water, moIes/liter ‘REFERENCES
the water-salt interface, moles/liter .1. Dommers, 0, B. and Jessen, F. W.: Feasibility of
es. = C at
Creating Spherical Cavities in Underground Sait
D = diffusivity constant, sq cm/sec Domes, The U, of Texas, Dept. of Petroleum
g = acceleration due to gravity, sq cm/sec 2 Engineering, Report to the U. S, Atomic Energy
Commission (Jan. 31, 1961).
mass of sak in a- volume of water of unit 2, Eckert, E, R, ,G. and Drake, R, M., Jr.: Heat and
gin/cc Mass ‘l%ozsfer, McGrsw-HilI Book CO, (19S9) 131,
salinity,
moles/lit”er “3; Re”rnson, D.’ R.:’ Model Stiniies” oj_ Techtii@es for
Dissolving Spherical Cauitiea in Massive Satt,, Ms
average maas of salt removed per unit surface Thesis; The U. of Texa,s (.Tan.. 1962).
area 4. Durie, R. W.: Mechanism of- tbe Di+solatfon oj
Salt in. the Fqrmation of Underground Salt Cavities, .
mass of salt removed per unit surface area
MS Thesis, The U. of Texas (June, 1962).,
at a point on the surface $ Wilke, C. R., ~ ~1,: qrree.convection Mass Transfer “”
.
time, sec at Vertical Pleteat~, Cbem. “ErIg. Prog, (1953) 663.
***
-. . .