Jökull - 01.12.1987, Page 39

(2) The CO, concentration in the parent water from
which the steam is derived is that given by the
C02-temperature function for equilibrated reser-
voirwaters by Arnórsson and Gunnlaugsson (1985).
(3) The N2 concentrations in the parent geothermal wa-
ter as well as in the cold shallow water are those of
non-thermal water in equilibrium with the atmo-
sphere.
(4) The C02 and N2 are unreactive in the upflow.
(5) Neither C02 nor N2 partition significantly into the
condensate/steam heated water.
These simplifying assumptions are discussed briefy be-
low.
It is difficult to provide evidence regarding the validity
of the first assumption of the boiling process in the
upflow. However, as discussed later, the low N2 concen-
trations in the fumarole steam at Krísuvík are taken to
be indicative of extensive boiling in the upflow, exten-
sive meaning that the water has boiled over a large
temperature (pressure) interval (>50°C) without being
separated from the steam phase.
Drillhole data were used to derive the temperature
function for C02 concentrations in geothermal reservoir
waters. They show some scatter which may be due to
several reasons as discussed by Arnórsson and Gunn-
laugsson (1985) including: a) departure from equilib-
rium; b) differences in the composition of minerals par-
ticipating in the equilibrium; c) imperfections in crystal
structures; d) errors in selected aquifer temperatures
and e) presence of equilibrium steam in the reservoir.
The scatter amounts on average to 0.27 logm CO, units.
For 280°C reservoir water such departure would give at
the most 9% condensation, if there was actually none.
The third assumption is based on the fact that geother-
rnal waters are dominantly meteoric in origin and it is
valid as long as the cold parent water has approached
fairly closely equilibrium with the atmosphere and mag-
matic contribution is not significant, or contribution
from other sources, and as long as degassing has not
occurred prior to boiling in the upflow. This assumption
finds support in drillhole data (Arnórsson, 1986).
N2 is the dominant nitrogen bearing species in Icelan-
dic geothermal systems, including the Krísuvík field.
Any reduction that might occur into NH3 or oxidation
into N03, will, therefore, not significantly affect the N2
concentrations, suggesting that N2 is effectively unreac-
tive.
Geothermal reservoir waters are generally very close
to being calcite saturated (Arnórsson, 1978; Arnórsson
etal., 1983b; Ellis and Mahon, 1977). Upon boiling they
become calcite supersaturated (Arnórsson, 1978) caus-
ing some removal of carbonate from the fluid, thus
lowering its C02 content. Total carbonate concentra-
tions in geothermal waters vastly exceed those of calci-
um, except for saline waters below some 150°C. There-
fore, the amount of carbonate removed from solution by
calcite precipitation is only a small fraction of the total
carbonate due to limited availability of calcium. Con-
densation of steam, whether by conductive heat loss or
mixing with cold water, produces an acid and calcite
undersaturated aqueous phase. Condensation does not,
therefore, lead to removal of C02 from the steam phase
through precipitation.
Both C02 and N2 are sparingly soluble in water. For a
system at 100°C and with 5% steam by weight 99.7% of
the mass of C02 occupies the steam phase and over
99.9% of the N2. At 250°C the same figures are 88% for
C02 and 98% for N2. It is, thus, evident that only very
extensive condensation at high temperatures will cause
significant partitioning of C02 into the condensate. Un-
der all conditions practically all the N2 occupies the
steam phase.
The following function describes the relation between
N2 concentrations in steam at atmospheric pressure and
the temperature of the parent water assuming adiabatic
boiling:
logN,c = 22.965 - 0.117235 • T +
2.05872 • 10'4 • T2 - 1.22958 • 10*7 • T:i (1)
where N2 is in mmoles per kg of steam and T in °K. The
subscript c denotes that the N2 concentration is a calcu-
lated value to distinguish it from measured concentra-
tion in the steam (N2 m). When deriving function (1) it
was assumed that the N2 content of the parent water was
0.71 mmoles/kg which corresponds with saturation at
5°C at PNj = 0.78 atm. For a parent water saturated with
N2 at 25°C (0.45 mmoles/kg) equation (1) becomes:
logN2 c = 22.700 - 0.116818 • T +
2.05013 ■ 10'4 • T2 - 1.22377 • 10“7 • T3 (la)
As C02 concentrations in the reservoir water are de-
termined by its temperature and the N2 concentrations
are taken to be constant, it follows that the C02/N2 ratio
in that water is a function of temperature only. The
following function describes the temperature depend-
ence:
t = 148.5 + 64.35 • Qcn +
5.239 • Qcn2 - 1.832 • Qcn3 (2)
where Qcn is log(CQ2/N2) in moles and t is in °C. The
37