Vol, 58

WILLIAMT. HOLSER

316

FUGACITY OF WATER AT HIGH TEMPERATURES AND PRESSURES'

BY WILLIAMT. HOLSER
Depzrtment of Geology, Cornell University, Ilhaca, New YorL
Receiised October 17, 1969

Fugacities for water in the single phase range up to 1000" and 2000 bars are calculated by graphical integration of recently
published p-FT data, and tabulated in terms of the fugacity coefficient. At high pressure and low temperature the fugacity
coefficients are slightly less than the approximate values predicted by Newton's empirical graphs.

Reactions involving water at high temperatures

and pressures are fundamentally important in geochemical processes, and are of increasing interest
in other fields. Consideration of equilibria in
such systems necessitates a knowledge of the
fugacity of mater.

Method
Accurate values of the fugacity of water may
be derived from p-v-T data published recently
by Kennedy.2 These data are based on measurements up t o lOOO", with an upper pressure limit
decreasing from 2500 bars at 600" to 800 bars a t
1000; values for higher pressures were extrapolated assuming constant [dp/bT],. For the
range to 460" and 350 bars, the previously published valuesS were taken by Kennedy as more
accurate. The fugacity, f, is defined4 with reference to the molal free energy in the standard
(ideal gas) state a t the same temperature, &
by the equation
RT 1nfT.p R , p - p+
(1)
whence

by substituting an equation of state7 valid for this

range
(5)

where

B T .=
~ BOT

and
BOT= 1.89
QT =

g ? ~=

(BOT?.))*

BIT

+ (BOTp)4

g2T

-jiji

- 2641.62 XT 1080s70/T

82.546

T -

0.21828

1.6246 X IO5

T2
1.2697 X lo5

giving

The first term suffices at high temperatures.

At 200 and 300" the isotherms meet the saturation line at 15.55 and 85.91 bars, respectively, so
that the above integration does not apply for the
range 0-100 bars. At the saturation line the
phase corresponding to the high pressure single
phase is the denser of the two, but its fugacity
must equal that of the lighter phase with which it
is in equilibrium. Equation 6 is therefore solved
for
the saturation pressure, and extended from
where v is the molal volume. Since Kennedy's
data have not been reduced to a,n equation of state, there by the graphical integration.
Results
the data were plotted on a large scale and inte
grated by planimeter along isotherms.
Fugacity coefficients calculated by the above
The results are most conveniently expressed6 in process are shown in Table I. Multiplication by
terms of the fugacity coefficient
the pressure gives fugacity. If desired, equation
(1) gives free energy in terms of the fugacity and
zJ =fh
(3)
the free energy in the standard state, taken from
as it changes less rapidly than f or F with p and T , standard tables.8
and is therefore more suitable for interpolation,
If activities corresponding to liquid state definiextrapolation or graphical representation.
tions are preferred, it may be recalled that unit
The mechanical integration is inconvenient for activity for a given temperature is defined as the
pressures lower than 100 bars (or perhaps 300 bars fugacity at one atmosphere pressure and the
at high temperatures). Base values for these low same temperat~re.~Above the critical temperapressures were calculated analytically from the ture the fugacity at such a low pressure is very
equation6
nearly equal to the pressure, as shown in the first
row of Table I, and the activity at higher pressure
RT In Y T , ~=
dp
(4)
is
- (C -

Lp

$)

(1) This work was supported in part by the Geological Society of

America and in part by the Office of Naval Research.
(2) G. C. Kennedy, Am. J . Sci., 848, 540 (1950).
(3) J.
Keenan
I. and F. G . Keyes, "Thermodynamic Properties of
Steam," John Wiley and Sons, Inc., New York, N . Y., 1936.
(4) G. N . Lewis and M. Randall, "Thermodynamics," McGrawRill Book Co., Inc., New York, N. Y., 1923, p, 190.
(5) M. A. Paul. "Principles of Chemical Thermodynamics," McGraw-Hill Book Co., Inc., New York, N. Y., 1951, p. 236. Thefugacity
coefficient is equivalent t o the "activity coefficient" of R. H. Newton,
Ind. Eno. Chem., 87, 302 (1935).
(6) M. A. Paul, ref 5, p. 237.

(7) F. G.Keyes, L. B. Smith and H.T . Gerry, Am. Aead. Ar1a Scd.,
70,319 (1936). Units for the Constants are retained from the original;
p is pressure in int. atm., T is in OK.
(8) F. D. Rossini and others, "Selected Values of Chemical Thermodynamic Properties," Nat. Bur. Stds., Washington, 1949-, Ser. 111,
Vol. 1. Free energy (referred to liquid water in equilibrium with vapor
at Oo) in the lower range of pressures and temperatures may be calculated directly from heat content and entropy data tabulated by J. H.
Keenan and F. G. Keyes, ref. 3.
(9) G. N. Lewis and Merle Randall, ref. 4, p. 256.
o is specific volume in cc./g..

FUGACITY
OF WATERAT HIGHTEMPERATURES
A N D PRESSURES

April, 1954

317

TABLE
I
FUGACITY
COEFFICIENT
OF WATER
Pressure,
bars

1
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400

1500
1600
1700
1800
1900
2000
* Two phase region-do

200

0.995

300

0.998

,134
.0703
.0497
,0392
.033 1
,0290
.026l
,0253
,0225
.0213
,0203
.0196

.701
.372
.259
.204
,172
.150
,135
,124
,115
.lo9
.lo3

.0190

.0956

,0991

.0928
,0185
.0181
,0900
,0887
.0178
,0872
,0176
,0174
,0860
,0173
.OS50
,0172
.OS43
not interpolate.

*
*

Temperature, O C .
600

400

500

700

800

900

1000

0.999

0.999

1.000

1.000

1.000

1.000

1.000

,874
.737
.602
,483
.412
.361
.325
.299
,281
.264
.251
.241
.233
.226
.220
,216
.211
.208
.205
.203

,923
.857
.764
,735
.655
,599
.555
,517
,472
.455
,436
.420
,400
,395
.385
.377
,370
,365
.360
* 356

,952
.909
,867
.830
,789
.751
,747
.693
,666
,642
.622

,969
.938

.980
,960
,941
.923
,906
.890
.881
.868
,861
.851
,842
.834
.827

,987
.975
.962
,950
.943
.931
.927
.919
,916
.910
.907
,904
.901
'899
,898
,897
,897
.896
.896
.896

,993
,985
.979
.963
.956
,951
,948
.947
,944

'

,542

.913
.886
,861
.838
.SI8
* 800
,783
.765
,750
,737
.725
.715
.706
,698
,691
,685

,535
.530

,680

,800

.676

.798

,604

.589
.577
.566
,556
.548

.821
,816

.811
.SO7
,803

Between 100 and 37A0, high density ("liquid")

water cannot be brought down to the 1 atm.
standard state without a phase change. If activity is to have any meaning in this range, it
must be in terms of the fugacity of hypothetical
liquid water at 1 atm. An approximate value
may be calculated by extrapolating the specific
volume of liquid water from the empirical formula
of Keenan and Keyedo

.944
.944

,945
,946
.948
,950
,952
,955
.957
.959
.962

probably a less satisfactory function than fugacity

above the normal boiling point of a fluid.
Discussion
The data of Table I do not all lie on smooth
curves. Measurements and calculations have been
double checked in all such cases. However, no
attempt was made to redraw the p-V isotherms to
give a smoother series of values. Some 7jf the
discrepancies may have originated in drawing the
ZI = 3.086 - 0.899017(374.1 - t)0.'47166 isotherm curves, and some may lie in the original
0.4(385 - t ) - ' . o ( p - 218.5) + 6 (8)
data. Both could be corrected by reduction to an
6 is a function of p and t, that was determined equation of state using least squares adjustment,
graphically by Keenan and Keyes. However, as has already been done for the lower temperatheir determinations are not available so that this tures and pressures. In plotting the p-V isofactor will have to be neglected here, although it is therms no deviations from the smooth curves were
probably insignificant except close to. the critical noticeable except a t lOOO", where the deviations
point. Integrating according to equation 2 from from the regular curve were somewhat more than
the stated precision of the measurements.
1 atm. to the saturation pressure p s gives
Newton has compared fugacity coefficients of
various gases by means of the law of correspondRT In fT.ps = [3.086 - 0.899017(374 - t)0.'471o8 +
fT, 1 atm.
ing states.I2 At low temperatures and high pres87.5(385 - t)-'.'](p, - 1) 0.2(385 - t)-'.' (psa - 1)
sures the values in Table I above are somewhat
(9)
less than the average of the several substances
In conjunction with the fugacities at the saturation used by Newton in constructing his curves. For
point already calculated above from equation 6, example, a t 400" and 2000 bars, Newton's curves
this gives fugacit,ies f 2 0 0 0 , I atm. = 13.62, and predict v = 0.24, compared with 0.20 calculated
(Table I). At high temperatures the calculated
f3000, 1 atm. = 67.69. These values are used in the
denominator of equation 7 to calculate activities a t values are a little higher than predicted. The
high pressure from the fugacity coefficients of the former discrepancy is not surprising in view of
first two columns of Table I. Lewis and Randall11 equation 5, which requires more than 3 indehave pointed out the compIicat>ionsarising from pendent constants for the equation of state, in
multiple definitions of activity. Because of the which case the law of corresponding states would
hypothetical character of its definitive standard not be expected to hold exactly. At the low
state and ambiguity of definition, activity is molal volume corresponding t o this temperature
and pressure, even equation 5 is inadequate as an
(IO) J. H. Keenan and F. G. Keyes, ref. 3. P . 20. Units for the
equation of state.'3
constants are retained from the original: G is specific volume in cc./g.,
'

p i s in int. atm., t is in " C .

(11) Reference 4, p. 257.

(12) R. H. Newton, ref. 5.

(13) F. G.Keyes, L. B . Smith, and H. T. Gerry, ref. 7.