The following text was automatically extracted from the image on this page using optical character recognition software:
PROCEEDINGS, Twenty-First Workshop on Geothermal Reservoir Engineering
Stanford University, Stanford, California, January 22-24, 1996
SGP-TR-151
MODELLING CHLORIDE AND CO2 CHEMISTRY AT THE
WAIRAKEI GEOTHERMAL FIELD, NEW ZEALAND
W.M. Kissling and S.P. White
Institute for Industrial Research and Development,
PO Box 31310 Lower Hutt, New Zealand.
M.J. O'Sullivan and D.P. Bullivant
Department of Engineering Science, University of Auckland,
Private Bag 92019 Auckland, New Zealand.
K.L. Brown
Institute for Geological and Nuclear Sciences,
Private Bag 92019 Auckland, New Zealand.
ABSTRACT
The chloride and CO2 chemistry at the Wairakei
geothermal field, New Zealand has been modelled us-
ing an extended version of the geothermal simulator
TOUGH2 which solves the equations for the transport
of reacting chemical species in multi-phase fluids. Re-
actions involving the speciation of aqueous CO2 to
H2CO3 and HCO~ are included in the model. Be-
cause CO2 speciation in water is pH dependent, a reac-
tion involving the most important weak acid buffer at
Wairakei (H4SiO4) has also been included. A 'Henry's
Law' reaction expresses the equilibrium between the
aqueous and vapour components of CO2. The chloride
is treated as a conservative, non-reacting species which
is present only in the liquid phase. Results from the
model are compared with measured chloride and CO2
data from Wairakei covering the period 1959 to 1987.
INTRODUCTION
In this paper we describe the first application of a chem-
ical transport model to a detailed, large scale simula-
tion model of a geothermal field. Our aim is to ver-
ify that some of the most important chemical processes
which take place in a geothermal environment (for ex-
ample boiling, dilution or deposition) do occur in the
model. We do this by comparing the modelled chem-
istry with observed chemical data. Having verified the
chemical model in this manner, we are able to pre-
dict future chemical changes that will take place in the
field under exploitation. As an example, it is possi-
ble to include calcite deposition about two-phase wells
in a reservoir model. This would allow determination
of production scenarios that minimise the effect of re-
duced permeability resulting from such deposition.
For this study we have chosen the Wairakei field in
New Zealand. This is a well studied geothermal field
which has been extensively modelled in the past. Pro-
duction of fluid started at Wairakei in the 1950's,
and chemical and other data has been collected con-
tinuously from that time. A good summary of this
data and early modelling studies of Wairakei is given
by O'Sullivan and McKibbin (1989). In this study
we use the latest in a series of 3-D Wairakei mod-
els (O'Sullivan and Bullivant (ECNZ, 1992)) with a
new chemical transport simulator to further extend this
model, and to check its validity against new chemical
data.
We have chosen to model the transport of the chlo-
ride ion (Cl-) and carbon dioxide (C02) in this study.
These species complement each other because they
each behave in a fundamentally different way when
subjected to the various processes, such as boiling and
dilution, which occur in a geothermal reservoir (see for
example, Henley et al., 1984). Consequently, mod-
elling Cl- and CO2 provides us with independent in-
formation on these processes, and can give new in-
sights into behaviour of the model.
Previous chemical modelling studies have been carried
out with both chloride and C02, although not, to the
authors' knowledge, simultaneously. The MULKOM
(Pruess, 1982) simulator has been extended by a num-
ber of workers (Mendrinos and O'Sullivan, 1990,
White and Kissling, 1992) to include the flow of chlo-
ride as a single non-reacting chemical species. In these
codes the chloride is assumed to be a conservative
217
