Jökull - 01.12.1987, Blaðsíða 29
Fig. 18. The conceptual model of the Leirbotnar and
Suðurhlíðar geothermal fields. — Hugmyndalíkan af
jarðhitasvœðinu við Kröflu.
(Fig. 18), 2) quantify the natural mass and the heat flow
in the reservoir, 3) verify the transmissivity values ob-
tained from the analysis of injection test data 4) estimate
the generating capacity of the field, and 5) predict on a
well-by-well basis the future field performance for al-
ternative field development plans including the rein-
jection of waste fluid.
The Krafla modeling has been described extensively
by Bödvarsson, et al. (1984a, b, c) and Pruess, et al.
(1984). Their results are summarized below. The natural
state model developed reproduced well the observed
pressure and temperature data for the field. The model
consisted of a vertical cross section which included both
Leirbotnar and Suðurhlíðar (Fig. 18). The simulation
model is in agreement with the assumption that the
reservoir system is controlled by two upflow zones, one
at Hveragil and the other close to the eastern border of
Suðurhlíðar. The lower reservoir in Leirbotnar and the
Suðurhlíðar one are two phase with average vapour
saturation of 10-20 % (volumetric) in the fracture sys-
tem. The porosity of the reservoir rock was assumed to
be 5%. The permeability of the reservoir is 1-4 millidar-
cy with an average of 2.0 md. The permeability seems to
be controlled by vertical fractures as better match was
obtained by allowing for higher vertical than horizontal
permeability. The permeability of the upflow channels
at Hveragil and Suðurhlíðar is estimated as 30 md.
Fluids from the upflow channel in Suðurhlíðar recharge
the reservoir at an estimated rate of 10 kg/s. The two
phase fluid mixture flows laterally along a highly perme-
able fracture zone at a depth of 1 km and mixes with the
upflow at Hveragil. The natural fluid flows are assumed
highest at Hveragil where 8 kg/s of steam are discharged
to the surface fumaroles. The remainder of the ascend-
ing fluid (13 kg/s) recharges the upper Leirbotnar reser-
voir.
Several models were developed to estimate the gener-
ating capacity of the reservoir and to predict its future
behaviour. In the most simple case a lumped parameter
model was used, but distributed parameter models and a
quasi three-dimensional model were also applied in or-
der to get more reliable estimates. All were based on the
natural state model. The results obtained indicated that
the generating capacities of Leirbotnar and Suðurhlíðar
are 30 and 20 MWe for 30 years, respectively. Future
exploitation will give rise to extended vapor zones in the
reservoir. Fluid enthalpy of production wells will rise
with time and reach dry steam conditions as the reser-
voir is depleted. The calculated flow rate decline for
individual wells is generally low or of the order of 1-4%
per year. Reinjection is predicted to change the future
behaviour of the wells. In the short term (few years) the
effect will be small, but when extended single phase
steam zones have formed (after 10 years) reinjection will
improve deliverabilities. Furthermore, it is possible that
reinjection could create single-phase liquid condition in
parts of the fracture system. This would increase the
fluid pressure considerably and increase the mass flow
rate of the production wells.
The quasi three-dimensional model used to predict
the future behaviour of the Krafla reservoir was cali-
brated by matching the flow histories (mass flow and
enthalpy) of wells that had been discharged for 1-6
years. This relatively short period influences the reliabil-
ity of the predictions of the model. So far all observa-
tions seem to be in agreement with the model, and for
instance pressure build-up tests in wells in Leirbotnar
and Suðurhlíðar in the summer of 1984 showed pressure
drawdown in the reservoir close to predicted values
(Sigurðsson, et al. 1985).
CONCLUSIONS
The Krafla geothermal area covers about 15 km2, but
only 3-4 km2 divided between three well fields have
been exploited so far. The production characteristics of
the three well fields differ widely. The transmissivity is
generally of the order 1-4 Dm. The enthalpy of the well
discharges is high in most parts of the fields thus render-
ing their exploitation efficient. The utilization of parts of
the area (northern Leirbotnar and the western flanks of
Mt. Krafla) has been greatly hampered by the intrusion
of magmatic gases into the geothermal system causing
blocking of wells by deposits and damages by corrosion.
There are signs now (1986) that such effects are dimin-
27