SPE
Society of PetroIetm EngIneers of AIME
SPE 12490
Downhole Density of Heavy Brines
by G. W. Krook: Dowell Division of Dow Chemical U.S.A.
T. D. Boyce, The Dqw Chemical Company
'Member SPE-AIME
Copyright 1984 Society of Petroleum Engineers of AIME
This paper was presented at the Formation Damage Control Symposium held in Bakersfield, CA, February 13-14, 1984. The material is subject to cor-
rectIon by the author. Permission to copy is restricted to an abstract of not more than 300 words. Write SPE, 6200 North Central Expressway, Drawer
64706, Dallas, Texas 75206 USA. Telex 730989 SPEDAL.
ABSTRACT
Heavy brine completion fluids decrease in density
wi th i ncreas i ng temperature and increase in dens ity
with increasing pressure. Pressure-volume-temperature
(P-V-T) analysis of various brine solutions at temper-
atures from 76° to 345°F and pressures of 0 to 22,000
psig has been used to examine the behavior- of brine
density under bottom-hole conditions. Information
obtai ned from these measurements allows the engi neer
to calculate more precisely the brine density at the
surface which will provide the desired hydrostatic
gradient for formation pressure control at downhole
temperatures and pressures.
INTRODUCTION
The effective use of heavy brines during comple-
tion and workover operations requires precise calcula-
tion of the hydrostatic pressure of the fluid column
at bottom-hole conditions. As with most fluids, these
brines expand with temperature and compress with
pressure. This affects the downhole density of heavy
brines used in such locations as the Gulf Coast, where
geothermal gradients can range from 1.2° to 2.2°F/l00
ft and hydrostatic pore pressure gradients can vary
from 0.433 to 1.0 ps i 1ft. Unfortunate ly, expansi-
bil i ty and compress i bil i ty data needed for these
bottom-hole pressure calculations at wellbore tempera-
tures and pressures are relatively limited, with
almost no published experimental data on heavier
brines and mixtures. Poor approximations can result
in severe losses of expensive fluid to the formation
or well control problems such as gas kicks.
In the absence of experimental data, many au-
thors 1-4 have used empi ri ca 1 equations to model the
density variations of various brines and other well-
bore fluids. Mathematical models have been recently
developed to successfully predict the downhole density
variance and hydrostatic pressure of a drilling fluid
column. 1 ,2 The calculations for these density models
for oil- and water-base muds were based on literature
References and illustrations at end of paper.
175
values for compressibility and expansibility of water,
sodium chloride solutions and oil. A modest amount of
density data modeling also exists for natural brines
and geothermal fluids, which contain sodium chloride
as the major electrolyte. 3,4 A P-V-T relationship has
a 1so been experimentally determi ned for vari ous con-
centrations of NaCl in solution to temperatures of
347°F and pressures to 4,978 pSi. s Recently, experi-
menta 1 measurements have been conducted on oil- and
water-base drilling fluids in the temperature and
pressure range of 70° to 400°F and 0 to 14,000 pSi. 6
Thi s paper exami nes for the fi rst time the be-
havior of brine density from laboratory P-V-T measure-
ments on NaCl, CaC1 2, NaBr, CaBr2' ZnBr jCaBr2 I CaC1 2 ,
and ZnBr/CaBr brine at pressures- of 0 to 22,000 pSTg
for constant 2temperatures of 76°, 198° and 345°F.
These laboratory studies show that the compressibility
and thermal expansibility of these fluids can vary
with brine composition, or more accurately, the total
salt concentration in solution. Furthermore, these
data have been used to develop a multiple regression
model that will accurately predict the change in brine
fluid density up to 345°F and 22,000 psig for salt
concentrations of 19% to 75% by wt. As a result,
optimum well control and cost control can be obtained
from the use of heavy brines during completion and
workover operations.
EXPERIMENTAl
P-V-T measurements were conducted using a Ruska
Model 2272 positive displacement pump designed to
operate up to a temperature of 350°F and a working
pressure of 25,000 psi. This pump is designed to
measure to within 0.01 cc the displacement of a piston
whi ch is forced into ali qui d chamber by a preci s i on
metering screw. Two of these pumps are shown in
series in a schematic drawing of the experimental
apparatus (see Figure 1).
The mercury charge pump with an 8,000-psi working
pressure and 250-cc pi ston di sp 1acement was used to
meter and measu re accurately the volume of test bri ne
transferred from the 500-cc reservoir to the P-V- T
displacement chamber. The P-V-T chamber was electri-
cally heated with a 750-w tape heater wrapped around
2 DOWNHOLE DENSITY OF HEAVY BRINES
the pump cylinder and covered with insulation and
metal sleeve. An LFE Series 230 Controller monitored
the cyl inder temperature from a built-in thermocouple
and maintained the temperature to within ±2.5°F. The
working pressure of the P-V-T displacement chamber was
25,000 psi. The pressure readings were made using an
Autoc 1 ave Engi neers Model AESC-20000-01-PE12 di gital
pressure indicator and transducer which were cali-
brated with a deadweight tester to obtain measurements
to within ±0.1%. The transducer was specially temper-
ature-calibrated to 350°F.
Measurement of the pump cylinder deformation at
the test temperatures and pressures was necessary in
order to correct the experimental data for the actual
change in volume of the bri nes at these conditions.
Therefore, the P-V-T chamber was first calibrated for
temperature and pressure effects against known proper-
ties of mercuryJ The procedure included measuring
the specific volume of mercury at 76°F and 2,000 psig
by metering with the charge pump a preCise volume of
fluid at 2,000 psig and then weighing its contents.
The mercury was then charged at 2,000 psig into the
evacuated P-V-T chamber with the piston fully dis-
SPE 12490
constant temperature. The results from these mea-
surements are presented in Figures 3 to 8. Note from
these fi gures that the temperature effects are IOOre
dominant on the change in brine specific volumes than
the pressure effects. A multiple linear regression
was used on the NaCl experimental data in Figure 3 and
was compared to volumetric properties obtained from
the literature on a 19% by wt NaCl solution (see Table
2). The values in these references were determined
from a regression analysis of compiled experimental
data from earlier studies. Experimental specific vol-
umes measured in this stu,dy were within ±0.5% of the
values determined by Rogers and Pitzer, and differ
from Potter and Brown's data by no more than 2.5%.
Two important thermodynamic ptoperties can be
derived from the experimental specific volume data.
They are the isobaric coefficient of volumetric expan-
s i on (a) and the i sotherma 1 coeffi ci ent of compressi-
bility (e). These are defined as
1 aV 1 av 1 !J.v
a=-(-) =-(-) - -
~

( - ) (2)
placed into the cylinder in order to measure the dead V aT P v aT P v !J.T P
volume around the piston and in the nipples between
the transducer and valves. This dead volume was 63.97 and
cc. The pi ston was then retracted by 200.00 cc (80%
of its total displacement) and the chamber was charged -1 aV -1 av -1 !J.v
again with mercury. Thus, the total working volume of e=-(-) =-(-) -(-) (3)
mercury at 2,000 psig was measured at 263.97 cc. V ap T v ap T v t§l T
Changes from this reference volume were then recorded
as the volume of mercury was slowly compressed in
2,000-psi increments to 22,000 psig at a constant Converting the specific volume to density and
temperatu re of 76°F. Then the volume occupi ed by evaluating the derivatives - (ap/aT)p and (apjap)T are
mercury was measured as mercury was slowly decom- quite useful for expressing the variation of density
pressed from 22,000 psig in similar increments. This as a function of temperature and pressure. These
cycle was then repeated. Therefore, the specific density changes are termed the brine density correc-
volume change with pressure at 76°F was calculated tion factors (A, B) and are defined as
from the average of four volumes measured at each
pressure. This procedure was repeated for tempera- ap !J.p
tures of 198° and 345°F. Experimental displacement A = - ( -) :: -( - ) (4)
data for mercury are shown in Figure 2 and are com- aT P !J.T P
pared to predicted values calculated from the litera-
ture. The difference is the pump correction volume as and
a function of temperature and pressure. The pump
correction volume was estimated by the following ap
equation. B ( -) (5)
ap T
Pump Correction Volume =
The coefficients a, A, e, and B have been calcu-
lated for each brine using Equations 2 through 5 and
the experimental data at the temperature and pressure
The specific volumes (cubic centimeters per gram) conditions shown in Tables 3 and 4. Note from the
of saline solutions of NaCl, CaCl , NaBr, CaBr, values of A and B that all of the brines would require
ZnBr /CaBr dCaC~, and ZnBr /CaBr 2 wefe measured as 2a approximately a 15,000-psi increase in pressure to
func'tion OT temperature arrd pre~sure by procedures counteract the thermal expansion from a 100°F increase
similar to the procedure for mercury. The pump cor- in temperature.
rection volume was subtracted from the measured volume
change for each bri ne at each temperature and pres- Since the brine density is a function of tempera-
sure. The compositions of the brine solutions are ture and pressure, Equations 4 and 5 can be combined
shown in Table 1. Between samples, the experimental to form the total differential expressed as
apparatus was cooled to room temperature, thoroughly
rinsed with water, blown dry, and evacuated under ap op
vacuum. dp = ( - ) dP + ( -) dT (6)
ap T aT P

RESULTS AND DISCUSSION or

In the test procedure, the specific volume of

each brine was measured as a function of pressure at a dp = BdP - AdT (7)
176
SPE 12490 G. KROOK AND T. BOYCE 3

The coefficients in Equation 7 are nearly con- V Volume (cc)

stant for many fluids over considerable ranges of
temperature and pressure. A simple approximation X Salt Concentration (% by wt)
woul d be to assume that A and B for each bri ne are
constant over the temperature and pressure range of a. = Isobaric Coefficient of Volumetric Expan-
interest. Thus, Equation 7 becomes sion (vol/voljOF x 10- 4 )

8 = Isothermal Coefficient of Compressibility

P - P B (P - P ) - A (T - T ) (8) (vol/vol/psi x 10- 6)
2 1 2 1 2 1
v = Specific Volume (cc/g)

Therefore, the density variance described in
Equation 8 could be approximated through a knowledge
of the density correction factors (see Tables 3 and
4). However, the information on density variance
would be limited to only the brine systems evaluated
in this study.
In order to more precisely predict this density
change and to include other brine compositions, an
interpolation formula was "developed from the pressure-
volume-temperature-concentration (P-V-T-X) relation to
fi t the experi menta 1 results. There were 204 data
poi nts used in the mul tip 1e regression ana lys is. The
resultant model accounted for 98.8% of the total
variation in density difference. The density differ-
ence for this model was defined as the change in
density from ambient conditions (i.e., 76°F and 0
psig) to the predicted density at temperature and
pressure. " Fi gure 9 is a graphi cal representation of
the density difference model for a 65% by wt salt
solution. Utilization of this model is shown by an
example problem in the Appendix.
CONCLUSIONS
1. A precise positive-displacement pump can be used
for the P-V-T analysis of the liquid phase of
brines once the pump deformation for temperature
and pressure has been calculated.
2. Experimental data collected for NaCl are in close
agreement with literature values.
3. An observation from eva 1uati ng the deri vat i ves
-(ap/aT)p and (ap/ap)T) of each brine system
shows tnat these bri nes waul d requi re approxi-
mately a 15,000-psi increase in pressure to
counteract the thermal expansion from a 100°F
increase in temperature.
4. An interpolation formula has been developed from
the P-V-T-X relation to fit the experimental
results. The multiple regression model accounted
for 98.8% of the total variance in density dif-
ference for salt concentration from 19% to 75% by
wt in a temperature range 76 0 to 345°F and pres-
sures up to 22,000 psi.
NOMENCLATURE
A Density Correction Factor for Temperature
(lb/gal/lOO°F)
p = Density (g/cc, multiply by 8.345 for
lb/gal)
ACKNOWLEDGMENTS
Many useful suggestions and criticisms from James
Adams, P-V-T consultant for Ruska Instrument Corpora-
tion, were deeply appreciated. The authors wish to
thank Chuck Peil for his assistance in gathering the
experimental data and Tim Rey for providing consulta-
tion in the mathematical modeling of the data. This
research was supported by the Specialty Chemicals
Department and Dowell Division of Dow Chemical U.S.A.
REFERENCES
1. Hoberock, L. L., Thomas, D. C., and Nickens, H.
V.: "Here's How Compressibility and Temperature
Affect Bottom-Hole Mud Pressure," Oil & Gas J.
(March 22, 1982).
2. Sorelle, R. R., Jardiolin, R. A., Buckley, P.,
and BarriOS, J. R.: "Mathematical Field Model
Predicts Downhole Density Changes in Static
Drilling Fluids," SPE 11118 (Sept. 26-29, 1982).
3. Potter, R. W., II and Brown, D. L.: "The Volu-
metric Properties of Aqueous Sodium Chloride
Solutions from 00 to 500°C at Pressures up to
2,000 Bars Based on a Regression of Available
Data in the Literature," U.S. Geological Survey
Bull. 1412-C (1977).
4. Rogers, P. S. Z. and Pitzer, K. S.: "Volumetric
Properties of Aqueous Sodium Chloride Solutions,"
J. Phys. Chem. Ref. Data (1982) II, (1).
5. Rowe, A. M., Jr. and Chou, J. C. S.: "Pressure-
Volume-Temperature-Concentration Relation of
Aqueous NaCl Solutions," J. Chem. Eng. Data
(1970) 15 (1).
6. McMordie, W. C., Jr., Bland, R. G., and Hauser,
J. M.: "Effect of Temperature and Pressure on
the Density of Drilling Fluids," SPE 11114 (Sept.
26-29, 1982).
7. National Research Council, International Critical
Tables of ttimerical Data, Physics, Chemistry and
Technology, Vol. 1, McGraw-Hill Book Co., New
York (1928).

B Density Correction Factor for Pressure APPENDIX

(lb/gal/1,000 psi)
Examp 1e Problem. Determi ne the hydrostatic
P Pressure (psi) pressure or bottom-hole pressure (P b ) exerted by a
brine completion fluid knowing the brine properties
T Temperature (OF) and well conditions below.
177
4 DOWNHOLE DENSITY OF HEAVY BRINES SPE 12490

16.0 lb/gal (Brine Density at Surface)

65% by wt (Salt Concentration)
0.832 psi/ft (Brine Pressure Gradient)
76°F (Surface Temperature)
330°F (Bottom-Hole Static Temperature)
17,000 ft (True Vertical Hole Depth)
1.49°F/I00 ft (Well Temperature Gradient)
So 1uti on. Hoberock, Thomas and Ni ckens 1 have
shown that the bottom-hole pressure Pb (psi) can be
computed from the sum of the pressure changes (lIP)
over small increments of depth in feet (60) from
surface to total depth with knowledge of the brine
density as a function of temperature and pressure
p(T,P). The following equation represents this calcu-
lation.

0.052 [p(T,P)] 6D (9)

Using the density difference model in Figure 9

for the 65% by wt salt concentration, the following
data table can be constructed to calculate the density
and subsequent pressure changes as a function of
depth.
Density
Depth Ten.,erature Difference peT .P) AI'
~ {OF} (lb/gal} {lb/gal} ~
0 76 0 16.0 0
2,000 106 -0.075 15.925 1,656
4,000 136 -0.150 15.850 1,648
6,000 166 -0.225 15.775 1,641
8,000 196 -0.300 15.700 1,633
0,000 225 -0.375 15.625 1,625
12,000 255 -0.450 15.550 1,617
14,000 285 -0.525 15.475 1,609
16,000 315 -0.600 15.400 1,602
17,000 330 -0.637 15.363 799

Therefore, Pb = L LlP = 13,830 psi.

Evaluating the bottom-hole pressure using the
traditional approach of constant density (i .e., pres-
sure and temperature effect cancel each other) results
in the following.
Pb = 0.052 (16.0) (17,000) = 14,144 psi
Note the difference of 314 psi from the actual
hydrostatic pressure could be significant in the
effect i ve use of heavy bri nes duri ng compl et i on and
workover operations.

178
 TABLE 1
PHYSICAL PROPERTIES OF HEAVY BRINES
EVALUATED IN THIS STUDY

Salt Concentration Density at 76°F (glee)

Brine (% by wt) o pSigt 2,000 psigtt

NaCl 19 1.138 1.1412

CaC1 2 37 1.374 1.3779
NaBr 45 1.497 1.4996
CaBr 2 54 1.715 1.7165
ZnBr 2 /CaBr 2/CaC1 2 65 1.920 1.9255
ZnBr /CaBr 2 75 2.310 2.3140

tMeasurement made by Mettler/Paar density meter DMA 35.

ttMeasurement made gravimetrically as described in experimental section.

TABLE 2
COMPARISON OF VOLUMETRIC PROPERTIES OF 19% BY WT
NaCl SOLUTION (cc/g)

Temperature Pressure Exp. Literature Values

(oF) (psi) Datat Potter &Brown tt Rogers & Pitzerttt

122 2,900 0.8864 0.8842 0.8840

122 8,700 0.8758 0.8703 0.8736
122 14,500 0.8652 0.8598 0.8642
122 21,750 0.8520 0.8489

212 2,900 0.9097 0.9050 0.9080

212 8,700 0.8992 0.8857 0.8962
212 14,500 0.8886 0.8741 0.8859
212 21,750 0.8753 0.8606

302 2,900 0.9331 0.9337 0.9376

302 8,700 0.9225 0.9083 0.9233
302 14,500 0.9120 0.8929 0.9115
302 21,750 0.8987 0.8772

tSpecific volumes calculated from the empirical equation;

(cc/g) = 0.859954 + 2.597412 x 10- 4T (OF) - 1.822673 x 10- 6 P(psi); R2 = 0.994.
ttValues obtained from Potter and Brown (1977).3
tttValues obtained from Rogers and Pitzer (1982).4

SPE12490
 TABLE 3
EXPANSIBILITY OF HEAVY BRINE FLUIDS AT
12.000 PSI FROM 76°F TO 198°F
Density a A
Brine (lb/gal} (Vol/Vol/oF x 10- 4 ) (lb/gal/l00°F)

NaCl 9.49 2.54 0.24

CaC1 2 11.46 2.39 0.27
NaBr 12.48 2.67 0.33
CaBr 2 14.30 2.33 0.33
ZnBr 2 /CaBr 2 /CaC1 2 16.01 2.27 0.36
ZnBr/CaBr 2 19.27 2.54 0.48

TABLE 4
COMPRESSIBILITY OF HEAVY BRINE FLUIDS AT
198°F FROM 2,000 TO 12.000 PSI
Density B B
JJrine (lb/gal} (Vol/Vol/~si x 1O-6} (lb/gal/l.000 ~si)

NaCl 9.49 1.98 0.019

CaC1 2 11.46 1.50 0.017
NaBr 12.48 1.67 0.021
CaBr 2 14.30 1.53 0.022
ZnBr 2 /CaBr 2 /CaC1 2 16.01 1.39 0.022
ZnBr/CaBr 2 19.27 1.64 0.031

SPE12L~jO
250-cc Electrically
Heated Displacement
Chamber (Ruska Model 2272) r----------------~rr===" Digital Pressure
I ndicator with
Transducer

Drain for
Density 50O-cc
Temperature Measurement Brine Brine Fluid
Controller at 2,000 psi Reservoir

Hg

25O-cc Hg Charge
Pump (Ruska Model 2250)

Drain

Figure 1. Schematic of P-V-T Apparatus.

Literature Data for Hg

--- Experimental Data for Hg

--
()
()
8

c-
Q)
E
....... ..
...
::::::::::::::::::::::::::: ........::::: '.;;;;;;: :::;; ;;;ij ~ ~ j j ~ ij j; j j ji:: :: :::: ::i:................... . .........................
........... ,

Q)
()
«S
c..
en
o o......~=~
c
-
o
en -2
a.. ~~~~
-42L--4~~6--~8--~10~~1~2~1~4--1~6~~18~~2~0~22
Pressure (psig x 103 )

Figure 2. Mercury Calibration of P-V-T Chamber.

SPE1249Q
 0.95 0.78
Oi 0.94
"- Oi 0.77
g 0.93
-
Q) 0.92
"-

-Eg 0.75
0.76
§ 0.91
o 0.90 ::J
>u 0.89
o 0.74
>
~
Q)
Q.
en
0.88_.
0.87
--.
--. __•
--.--.--.__•
-
.-u
.u-
Q)
Q.
0.73
0.72
0.86 760F en 0.71
0.85 ---.
--.~
0.840 2 4 6 8 10 12 14 16 18 20 22 0.700 2 4 6 8 10 12 14 16 18 20 22
3
Pressure (psig x 10 ) Pressure (psig x 103 )

Figure 3. Specific Volume of a 19% by wt NaCl Brine Figure 4. Specific Volume of a 37% by wt CaC1 Brine at
at Elevated Temperatures and Pressures. Elevated Temperatures and Pressures.2

0.73
Oi 0.72 .---.--.---. -
0)
"-
0.63

g 0.62
"-

-gE 0.71 ---.---. - . -E

0.70 .-.-.
---.----345°F ::J
.g
0.61

0.60"
0 ____
~ 0.69 0-0_
0 ____ 0 ___

> 0.68 °-0- 0 - - - 0 _ °_0- 0 _ 0 198°F

u 198 ° F :E 0.59'"
°-0-°-0 -°-0-°_0
-0

0.57

0.640 2 4 6 8 10 12 14 16 18 20 22 0.560 2 4 6 8 10 12 14 16 18 20 22
Pressure (psig x 103 ) Pressure (psig x 103 )

Figure 5. Specific Volume of a 45% by wt NaBr Brine at Figure 6. Specific Volume of a 54% by wt CaBr Brine at
Elevated Temperatures and Pressures. Elevated Temperatures and Pressures~

SPE1249Q
 0.56 0.47
-8
----
Cl

CD
E 0.54
0.55 .----.---.
----.----.---~
.---.---.-.-.
345°F
-8
----
Cl

CD
0.46

:::J ~ 0.45
o
> 0.53 o
>
-g
()

c..
0.52 -() 0.44

en 0.51

0.50 0 2 4 6 8 10 12 14 16 18 20 22 0.420 2 4 6 8 10 12 14 16 18 20 22
Pressure (psig x 10
3
) Pressure (psig x 103 )
Figure 7. Specific Volume of a ZnBr 2 /CaBr 2 /CaCl Brine Figure 8. Specific Volume of a ZnBr 2 /CaBr Brine
(65% by wt Total Salt Concentrarion) at 4
(75% by wt Total Salt Concentratlon) at
Elevated Temperatures and Pressures. Elevated Temperatures and Pressures.

---.....
CD
:::J
CIJ
CIJ
CD
.....
c..

Temperature (OF)

NOTE: Lines are spaced at 0.05 Ib/gal

increments of density change; where
/:l.p = p (T,P) - p (76° F, 0 psig)
Figure 9. Density Difference Model for a 65% by wt Salt Solution.

SPE12490