Renewable Energy 86 (2016) 1295e1307

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Application of diffuse gas flux measurements and soil gas analysis to

geothermal exploration and environmental monitoring: Example
from the Reykjanes geothermal field, SW Iceland
Thra


n b, c, *, Finnbogi Oskarsson

inn Fridriksson a, 1, Eleazar Padro a
rez b, c
, Nemesio M. Pe
a
Iceland GeoSurvey, Grensa
svegur 9, 108 Reykjavík, Iceland
b
gico y de Energías Renovables, 38611 Granadilla de Abona, Santa Cruz de Tenerife, Canary Islands,
Environmental Research Division, Instituto Tecnolo
Spain
c
gico de Canarias (INVOLCAN), 38400 Puerto de la Cruz, Tenerife, Canary Islands, Spain
Instituto Volcanolo

a r t i c l e i n f o a b s t r a c t

Article history: Soil gas geochemistry and diffuse CO2 flux from soil, together with gas concentrations in 7 production
Received 20 June 2015 wells and 10 steam vents in the Reykjanes geothermal field (Iceland), were studied with the aim of: 1)
Accepted 17 September 2015 studying the potential use of the CO2/He ratio in soil gas to constrain CO2 retention in calcite in the
bedrock; and 2) quantifying the effect of 100 MWe power production from the Reykjanes geothermal
Keywords: reservoir on surface activity. The presence of biogenic CO2 in the soil gas prevented the use of the CO2/He
Geothermal energy
ratio for the quantification of CO2 retention in calcite. The CO2/He ratio in about 20 soil gas samples was,
CO2 retention
however, low and consistent with CO2 retention. Our results suggest that detailed soil gas and soil
Soil gas
Diffuse emissions
temperature mapping may provide excellent information for defining drilling targets. The surface activity
Reykjanes geothermal field in the Reykjanes geothermal field increased noticeably after the commissioning of the 100 MWe Rey-
kjanes Power Plant in 2006. When increased steam flow from steam vents is taken into account the total
increase in surface activity in Reykjanes between 2004 and 2007 is of the order of 50%.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction
Geochemical exploration studies of geothermal systems are
traditionally focused on determining and interpreting the compo-
sition of geothermal fluids emerging from the geothermal reservoir.
The chemical composition of geothermal fluids, such as water,
steam, and dry gas, are used to constrain parameters such a
reservoir temperature, type of water, salinity, and extent of fluid-
erock interactions (e.g. Refs. [21,8,30]. The geochemical methods,
most widely used in geothermal exploration, have also been found
to be very useful for monitoring producing geothermal reservoirs
(e.g. Refs. [6,35]. In recent years soil gas surveys have become
increasingly common in geoscientific studies. Among the objectives
of these studies are the quantification of gas and heat flux from
volcanic and geothermal systems [13,14,27] and the identification
* Corresponding author. Present address: Environmental Research Division,

gico y de Energías Renovables, 38611 Granadilla de Abona, Santa
Instituto Tecnolo
Cruz de Tenerife, Canary Islands, Spain.

n).
E-mail address: eleazar@iter.es (E. Padro
1
Present address: The World Bank, 1818 H St NW, Washington, DC 20433, USA.
and delineation of active faults [16,49,58]. Both topics are of central
importance to investigations of geothermal resources.
In this paper we present results of a soil gas study carried out in
the Reykjanes geothermal field, SW-Iceland (see Fig. 1) involving
both determinations of the concentrations of soil gases (He and
CO2) and the gas flux through the soil (CO2). The objectives of this
study were 1) to explore and compare the applicability of soil gas
analyses and CO2 flux measurements for geothermal exploration, 2)
to quantify fractionation of CO2 relative to He in steam ascending
from the reservoir with the purpose of constraining the amount of
geothermal CO2 retained in the bedrock and 3) to quantify the
changes in soil gas emission from the Reykjanes field after the
commissioning of the 100 MWe Reykjanes power plant in 2006.
The Reykjanes geothermal field is an ideal site for this study.
First of all the areal extent of the active part of the field is relatively
small and it is thus manageable to map the soil gas composition and
diffuse soil degassing in detail. Previous studies have revealed that
well defined diffuse degassing structures exist within the active
part of the field [27] and the concentration and spatial distribution
of geothermal CO2 in the bedrock has also been quantified [61]. The
lessons learned from this study in terms of the comparison of the

http://dx.doi.org/10.1016/j.renene.2015.09.034
0960-1481/© 2015 Elsevier Ltd. All rights reserved.
1296 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307
Fig. 1. Location of the study area in the Reykjanes Geothermal Field, SW Iceland. Local roads are shown in gray. Geothermal exploration and production wells are shown with
crossed circles well tracks are indicated by black curves.
applicability and efficiency of the different methods for geothermal
exploration are valid for other high temperature geothermal sys-
tems. However, the quality of the actual results obtained by the
different method may differ strongly from one field to another.
Degassing through soil will vary strongly from one field to another
depending on a number of factors, including the boiling processes
in the system, gas content of the deep fluid, dimensions of an
overlying groundwater reservoir and soil type. If degassing activity
is limited none of the soil gas methods will provide useful infor-
mation. The results of this study do, nevertheless, provide valuable
comparison of soil gas methods for geothermal exploration.
2. Settings
The Reykjanes geothermal field is located within the Reykjanes
volcanic system [38] at the south western tip of the Reykjanes
peninsula (Fig. 1). This volcanic system is the westernmost system
in Iceland's western neovolcanic zone, which is the subaerial
continuation of the Reykjanes Ridge. The Reykjanes geothermal
system is one of >20 high temperature geothermal systems located
within the Icelandic volcanic zones [5,9].
2.1. Surface features
The Reykjanes volcanic system is characterized by oblique
extensional tectonics [17,20] and episodic fissure eruption volca-
nism [40,55]. The two most recent volcanic episodes occurred in
the late twelfth and the early thirteenth century and between 1900
and 2100 years ago [54,55]. The landscape is dominated by recent,
low-relief lava flows and volcanic crater rows but a few Pleistocene
hyaloclastite ridges protrude through the younger Holocene lavas.
The vast majority of faults in the Reykjanes field are normal faults
striking SW (030
e060
[17,18]). Clifton and Schliche [17] also
describe a more NeS trending right lateral oblique-slip fault and
this orientation is close to the strike of the diffuse degassing
structures in the Reykjanes field described by Fridriksson et al. [27].
Geothermal manifestations in the Reykjanes field include steam
vents, hot ground, mud pits, fracture emanating warm and water-
saturated air, and fossil surface alteration [27]. The aerial extent
of geothermal manifestations at Reykjanes is approximately 2 km2
[50]. Before commissioning of the 100 MWe Reykjanes Power Plant
in 2006 most of the current surface activity was concentrated in an
area of approximately >0.5 km2 south of the Gray Lagoon [27];
Fig. 1) but since then hot ground activity has expanded significantly,
particularly within the well field north of the lagoon.
2.2. Subsurface features
Thirty five exploration, reinjection and production wells have
been drilled at Reykjanes to date (including forked legs). The
stratigraphic structure is relatively simple; Holocene lavas extend
down to a ~150 m, subglacial/submarine hyaloclastite formations
(tuffs, breccias as tuffaceous sediments) and marine sediments
dominate the stratigraphy to ~1000 m and pillow basalts and
basaltic intrusives dominate below ~1200 m depth [56,24,25,44].
The highly permeable Holocene lavas and the uppermost part of
the hyaloclastites (down to about 500 m) host a cold groundwater
system and the outflowing of geothermal fluids [24] except in the
steam vent field south of the lagoon (Gunnuhver area) where hot
fluids ascend toward the surface. The sediments below 400e500 m
form a cap-rock for the geothermal system preventing the inflow of
relatively cold saline groundwater into the deeper part of the
system.
Productive aquifers, with temperatures of 275e315
C, are
encountered from ~1000 to 2500 m depth. Temperature profiles
show that temperature increases only slightly with depth through
the productive reservoir indicating efficient convection. The
geothermal solutions are seawater that has been modified by re-
actions with the host rocks at reservoir temperatures [10,4,36]. The
concentration of dissolved CO2 in the geothermal solution is of the
order 900e2000 mg/kg [26] and it is considered to be derived from
degassing of crystalizing dike intrusions and partly from leaching of
 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307
volcanic and intrusive host rocks [4].
The secondary mineralogy is reflects the temperature distribu-
tion in the reservoir [25,44]. The characteristic alteration zones, in
order of increasing alteration grade, are: mixed-layer smectite-
chlorite, chlorite, mixed layer chlorite-illite, epidote, and actinolite
zones [42,25,44]. Calcite distribution has been quantified by carbon
analyses of drill cuttings [48,61]. These studies have shown that
calcite has accumulated in the uppermost 500 m of the crust, i.e.
above the convective geothermal reservoir, where it constitutes on
average ~4% of the mass of the rock, while this mineral is generally
absent or present in only minute amounts at the reservoir level.
3. Methods
A survey of CO2 flux and CO2 and He concentrations in soil gas
was carried out in the Reykjanes geothermal field in June and July
2007. The survey area covered the most significant surface
geothermal activity of the Reykjanes field. Soil temperature and
CO2 flux was measured in the same area in 2004 [27] also in June
and July and in similar weather conditions. The extent of the area
covered by soil gas measurements was about 0.25 km2, whereas the
CO2 flux through soil and soil temperature was measured on an
area of 0.3 km2. Measurements were carried out on a grid with
approximately 25 by 25 m spacing. Soil gas samples were collected
from a total of 305 points and CO2 flux was measured at all these
locations. In addition, some 22 CO2 flux measurements were car-
ried out in the most active part of the geothermal field where soil
gas sampling was not possible and 83 CO2 flux measurements were
carried out at the SE border of the area. The location of each sam-
pling/measurement point was recorded with a hand-held GPS unit
with a reported accuracy of 3 m and the soil temperature was
measured at 15 and 45 cm depth. Furthermore, complete steam
samples were collected from 7 production wells and 10 steam vents
for gas analysis and samples for He analysis were collected from 6
additional vents.
3.1. Soil gas analysis
At each sampling site, soil gas samples were collected at 40 cm
depth using a stainless steel probe. The probe is a stainless-steel
tube fitted with a permeable ring just above the pointy bottom
end and a rubber septum on the top. A steel rod was used to pre-
pare a hole for the probe which was subsequently inserted into the
ground and the soil around pressed towards it in order to prevent
atmospheric contamination along the probe. Once the probe had
been inserted gas was drawn from it with a 60 ml syringe through
the septum. The probe was flushed by drawing at least three times
60 ml through it before the gas sample was inserted into vacutainer
tubes [37]. Vacutainer tubes are glass test tubes filled with a dilute
solution of HCl and CuCl2 and fitted with a septum. HCl and CuCl2
were added to avoid CO2 dissolution in the water and to prevent
bacterial growth which could distort the CO2 concentration in the
head-space. When a gas sample is introduced to a vacutainer from
the syringe the needle is inserted through the septum, which faces
down. Another needle is inserted through the septum in order to
allow liquid to escape when the gas sample is introduced. The
vacutainer is subsequently kept with the septum facing down until
the sample is analyzed.
The soil gas samples were analyzed for He and CO2 concentra-
tion within 5e6 h by means of quadrupole mass spectrometer
(QMS; Pfeiffer Omnistar 422). Atmospheric gas and several CO2
standards were used periodically to calibrate the instrument. The
gas samples were injected into the QMS by a needle totally inserted
into the vacutainer and connected to a 1 m long steel capillary,
thermostated at 150
C to prevent condensation of water. The
1297
inflow rate into the spectrometer ranged between 1 and 2 cm3/min.
The estimated accuracy of the helium concentration as determined
by the instrument was ±300 ppb. The algorithm used for the
quantitative analysis module of the QMS was presented by Gander
[29].
3.2. Sampling and analysis of steam from steam vents and
production wells
Samples for He analysis were collected from 16 steam vents and
samples for complete gas analysis were collected from 10 of these.
Complete gas samples were collected from 7 production wells. The
selected steam vents are distributed all over the relatively small
geothermal field. Steam vent sample locations and He concentra-
tions in the steam are indicated in Fig. 2. Samples were collected
from all but two production wells accessible at the time of the
study. The steam vents and production wells sampled are thus
considered to be representative of the system.
Samples were collected from steam vents using a plastic funnel
connected to a 100 diameter titanium pipe. Care was taken to cover
the funnel with clay or soil to prevent atmospheric contamination
during sampling. Steam samples from production wells were
collected using a Webre-separator.
Samples for He analysis from steam vents and production wells
were collected through a silicon rubber hose that was connected to
the titanium pipe of the Webre-separator on one end and a three-
way valve on the other. A syringe was connected to another port of
the three way valve and the third was used for flushing the system
with steam. When the system had been flushed properly the steam
was allowed to pass into the syringe where it condensed and the
non-condensable gases accumulated. The gas fraction was then
passed to a water filled vacutainer for subsequent QMS analyses as
described in Section 3.1 above.
Steam samples for analysis of CO2, H2S, N2, H2, Ar, O2, and CH4
were collected onto evacuated, double-port Giggenbach bottles
containing 40% w/w NaOH solution (see e.g. Ref. [7]. Upon sampling
the water vapor condensed and CO2 and H2S dissolved in the
caustic soda solution. The head space gases were analyzed with a
PerkineElmer Arnel 4019 Gas Chromatograph. CO2 and H2S were
analyzed for by titration; CO2 by potentiometric titration (pH
titration) and H2S by HgAc2 titration with dithizone indicator [1].
3.3. CO2 flux measurements
The CO2 flux was measured directly with the accumulation
chamber method [13,51] using a closed system CO2 flux meter from
West Systems (Pisa, Italy) equipped with a LICOR LI-820 single-
path, dual wavelength, non-dispersive infrared gas analyzer. All
flux measurements were made using a chamber with 3.06 103 m3
total internal volume and 3.14 102 m2 basal area. CO2 flux mea-
surements were only conducted on dry ground to avoid non-
representative gas flow conditions due to soil saturation (see Ref.
[34]. When the CO2 flux was measured the chamber was pressed to
the ground firmly enough to isolate the internal volume of the
chamber from the outside, taking care not to disturb the surface
during measurement as this can temporarily result in an anoma-
lously high gas flux from the soil.
3.4. Data processing by sequential Gaussian simulations
Sequential Gaussian simulations (sGs) are now commonly used
to process results of soil gas surveys [12,45]. The data set is used to
generate a number of equiprobable realizations of the simulated
parameter on a grid defined by the user. Each realization is
consistent with the statistical properties of the data set. In this
1298 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307

Fig. 2. Soil temperature at 15 cm depth. Locations of sampled steam vents are indicated by star-shaped symbols and the He concentration in samples from the respective vents, in
ppmV is shown. Red circles with black numbers show well head locations. (For interpretation of the references to colour in this figure legend, the reader is referred to the web
version of this article.)

study we used the sGs algorithm of the sgsim code by Deutsch and
Journel [19] to process the He and CO2 concentration in the soil gas
and the CO2 flux. Here the results of sGs analysis of the data sets are
presented as E-type maps that depict the “expected” value at any
location computed as an average value of the simulated parameter
for each cell of the model grid for all realizations. The variogram
and simulation parameters for the soil CO2 and He concentrations
and CO2 flux data are summarized in Table 1. Note that the vario-
gram parameters shown in Table 1 refer to normal scored data.
4. Results
4.1. Soil gas and soil temperature
for the atmospheric concentration all samples below 5.72 ppm are
interpreted as containing only atmospheric He. According to this
criterion a total of 200 out of 305 soil gas samples contained at-
mospheric He only.
CO2 concentrations in soil gas ranged from 0.033 to 91.2 vol%;
the median and mean values were 0.44 and 6.39 vol%, respectively.
Most soil gas samples had CO2 concentrations distinctly elevated
relative to atmospheric concentrations. The CO2 concentration in 4
samples only was within 10% of the atmospheric value in 2007
(0.0384 vol% [46]; and that of 23 samples within 50% of that value.
An E-type map depicting the results of 100 sequential Gaussian
simulations of the He and CO2 concentrations in soil gas are shown
in Figs. 3 and 4.

Soil temperatures in the study area ranged from near ambient 4.2. Gas composition of steam from production wells and steam
(8e15
C) to boiling water temperature. The temperature at 15 cm vents
depth ranged from 6.8 to 100
C and at 45 cm depth it ranged from
11.9 to 100
C. Fig. 2 shows a soil temperature map of the study area The gas composition of steam from production wells and steam
depicting the temperature at 15 cm depth. The map was generated vents is shown in Table 2 as vol% of dry gas except the He con-
by simple Kriging method. A corresponding map based on tem- centration, which is shown in the units ppmV. In all cases CO2 is the
peratures at 45 cm depth was essentially the same as the map in dominant gas, ranging between 89.8 and 94.3 vol% in the well
Fig. 2 except for slightly higher temperatures. samples and from 90.0 to 97.4 vol% in the steam vent samples. The
He concentrations in soil gas ranged from 4.8 to 118 ppm V; the other significant gases are H2S, N2, and H2. The H2S concentration
median and mean values were 5.46 and 8.07 ppm V, respectively. ranges between 3.4 and 5.7 vol% in the steam from the wells but
The lowest values are about 8% below the atmospheric concen- shows more variation in the vent steam, ranging from 0.1 to 5.8 vol
tration of He (5.2 ppm V; [47]. Taking ±10% as the uncertainty limit %. N2 concentrations range between 1.5 and 3.4 vol% in well steam
and 0.7 and 5.1 vol% in vents. H2 concentrations range from 0.14% to
1.25% in the well steam and from 0.23% to 1.79% in the steam from
Table 1 the vents. He concentrations in the well steam range from 35.4 to
Summary of variogram and sGs parameters. 46.3 ppm V and in steam from the vents it ranges from 17.9 to
44.1 ppm V. Oxygen was measured in all steam samples and was in
Parameter Variogram (for normal scored data) Simulation information
all cases found to be below the detection limit of the instrument
Model Nugget effect Range Cell size Realizations
(0.0005%), suggesting that the samples were not affected by at-
2
Soil CO2 Spherical 0.42 135 m 4 by 4 m 100 mospheric contamination during sampling. Absence of oxygen may
Soil He Spherical 0.42 135 m 4 by 4 m2 100 not conclusively prove that the samples are unaffected by atmo-
CO2 flux Spherical 0.22 200 m 2 by 2 m2 100
sphere. In fact, steam vent samples 2 and 5 show elevated
 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307 1299

Fig. 3. He concentration in soil gas. Red circles with black numbers show well head locations. (For interpretation of the references to colour in this figure legend, the reader is
referred to the web version of this article.)

Fig. 4. CO2 concentration in soil gas. Red circles with black numbers show well head locations. (For interpretation of the references to colour in this figure legend, the reader is
referred to the web version of this article.)

concentrations of N2 and Ar and H2S depletion which may indicate
some atmospheric contamination in those samples. The atmo-
spheric signature of these samples may result from mixing of
geothermal steam with atmosphere in the subsurface and subse-
quent reaction of O2 with H2S.
The total gas concentration of the steam from production wells
ranges from 151 to 253 mmol/kg at sampling conditions. The
samples were collected at 25.5 to 44 bar-g pressure and if the gas
concentrations are corrected to account for boiling to atmospheric
conditions the resulting total gas concentrations are 47e93 mmol/
kg of steam. The total gas concentration in steam from steam vents
is higher and has a wider range; from 148 to 687 mmol/kg. Of the
ten steam vent samples listed in Table 2 six exhibit a very narrow
range of total gas concentrations, i.e. between 148 and 170 mmol/
kg. This concentration range is very similar to the total gas con-
centrations in steam from production wells at sampling conditions.
There is a systematic pattern between the total gas concentration
and the relative concentrations of the individual gases. The six vent
1300 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307

Table 2
Gas concentrations in steam from production wells and fumaroles.

Well id Sample id Steam fraction Gas content at sampling Gas content CO2 H2S N2 H2 Ar CH4 He
at sampling conditions at 1 bar
conditions
mmol/kg steam mmol/kg steam vol% vol% vol% vol% vol% vol% ppmV

Well samples
RN-14b 20070185 0.17 180 73 91.7 4.19 3.05 0.973 0.054 0.016 40.2
RN-15 20070187 0.15 233 93 95.6 2.60 1.57 0.143 0.031 0.021 46.3
RN-18 20070186 0.16 151 63 93.8 3.51 2.03 0.618 0.040 0.017 35.8
RN-19 20070188 0.10 253 74 93.9 3.71 1.55 0.828 0.030 0.012 40.8
RN-22 20070182 0.13 145 47 89.8 5.66 3.17 1.249 0.058 0.018 38.4
RN-23 20070181 0.10 191 50 90.8 4.88 3.38 0.825 0.054 0.027 36.9
RN-24 20070183 0.13 179 64 94.3 3.44 1.92 0.308 0.033 0.015 35.4
Steam vent samples
G0701 20070155 148 91.6 4.96 1.63 1.786 0.023 0.019 44.1
G0703 20070167 687 97.4 0.06 2.30 0.227 0.039 0.015 17.9
G0704 20070168 169 92.0 4.69 2.04 1.255 0.030 0.019 38.8
G0707 20070171 170 91.6 5.36 1.64 1.383 0.022 0.021 41.0
G0709 20070173 162 90.0 3.82 5.09 1.016 0.067 0.020 39.5
G0711 20070175 153 91.4 5.78 1.50 1.314 0.023 0.019 40.7
G0712 20070176 153 91.3 5.58 1.63 1.438 0.025 0.019 39.0
G0713 20070177 367 96.8 2.16 0.69 0.299 0.012 0.015 24.3
G0715 20070179 346 94.0 3.65 0.75 1.525 0.011 0.017 20.0
G0716 20070180 412 96.6 1.93 0.66 0.801 0.012 0.019 35.2

samples whose total gas concentrations are similar to those of the
production wells are also characterized by similar relative con-
centrations of individual gases. For instance, the CO2 concentra-
tions of these samples range from 90.0 to 91.6%, whereas the CO2
concentrations in the four samples whose higher total gas con-
centrations range from 94.0 to 97.4%. All other gases show the in-
verse trend, i.e. their relative concentrations are higher in the six
low total gas samples and lower in the four high total gas samples.
For He the relative concentrations in the low total gas samples are
very similar to its concentrations in the well steam, i.e. between
38.8 and 44.1 ppm V, whereas in the four high total gas samples the
He concentrations range from 17.9 to 35.2 ppm V. The lowest dry
gas concentration of He is found in the sample whose total gas
concentration is the highest. In other words, it appears that the four
high total gas steam vent samples contain excess CO2 relative to
those from the other steam vent samples and production wells.
The spatial distribution of the two groups of steam vent samples
(high and low total gas samples) seems to be strongly correlated to
the intensity of the surface activity. The low total gas steam vent
samples, which are characterized by relatively high and consistent
He concentrations were all collected in the area of most intense
steam vent activity. The high total gas steam vent samples were, on
the other hand, all collected from vents that are located on the
peripheries of the surface activity. This can be seen in Fig. 2 that
shows the locations of the steam vents as stars and the He con-
centration in the steam superimposed on the soil temperature map.
The high He steam samples are clustered in the area where the
highest soil temperature is observed whereas the low He samples
are collected from steam vents where the soil temperature is more
moderate. However, it should be iterated that all the steam samples
that are considered in this study represent “proper” geothermal
steam with an exit temperature of about 100
C at the vent.
Interpretation of the gas chemistry of the deep reservoir liquid
in the Reykjanes system is beyond the scope of this communication.
However, in order to facilitate the interpretation of the analytical
results in Table 2 we list the computed steam fractions at collection
in the table along with the analytical results. The steam fraction was
computed from the observed sampling pressure and a reference
temperature for each well assuming that the enthalpy of the total
discharge is equal to the enthalpy of saturated liquid at the refer-
ence temperature. See Ref. [8] for further description of the calcu-
lation of deep liquid composition.
4.3. CO2 flux through soil
Fig. 5 shows the sGs results for CO2 flux through soil in 2007. The
CO2 flux ranged from 0.2 to 2060 g/m2/day; at roughly half the
points (143) the CO2 flux was <20 g/m2/day. Fridriksson et al. [27]
reported that only 5% of the non-geothermal flux values measured
in the 2004 campaign in Reykjanes were above 10 g/m2/day. As a
result 20 g/m2/day can be taken to be close to the maximum CO2
flux value for vegetated soil without a geothermal CO2 contribution
in the Reykjanes area. The 2007 CO2 flux anomaly shown in Fig. 5 is
largely consistent with the 2004 CO2 flux anomaly [27]; see Fig. 10)
with some exceptions. In 2007 the intensity of the CO2 flux in the SE
part of the study area had increased substantially relative to 2004
and the anomaly had also expanded spatially in that area. The CO2
flux in the NW part of the study area had, on the other hand,
decreased somewhat since 2004. The total CO2 flux from the study
area in 2007 was 18.5 t/day, compared to 13.5 t/day in 2004 [27].
The standard deviation of 100 realizations of the 2007 was 2.2 tons/
day but a corresponding value for 300 realizations of the 2004 data
was 1.7 ton/day. The changes in the magnitude of the CO2 flux and
the topography of the CO2 flux anomaly between 2004 and 2007
have been related to the commissioning of the 100 MWe Reykjanes
Power Plant in 2006 and will be discussed further in Section 5.5
below.
5. Discussion
5.1. CO2 and He in soil gas, steam vents and production wells
Among the objectives of this study are the definition of the
sources of CO2 and He of soil gases in geothermal areas, and of the
processes affecting the concentration of these two species in the
soil gas. The first step in this analysis is to identify the potential end
member components whose mixture determines the soil gas
composition. Two obvious endmembers can be identified:
geothermal gas and atmospheric gas. However, as will be discussed
in the sections below, the composition of the soil gas samples in the
Reykjanes field cannot be explained solely as a simple mixture of
these two endmembers.
Fig. 6 depicts the concentration of CO2 as a function of He
concentration in soil gas as red circles. Green and yellow circles
represent the CO2 and He concentrations in steam from production
 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307 1301

Fig. 5. CO2 flux through soil. Red circles with black numbers show well head locations. (For interpretation of the references to colour in this figure legend, the reader is referred to
the web version of this article.)

Fig. 6. CO2 concentration as a function of He concentration in soil gas samples and steam samples from steam vents and production wells. Also shown are hypothetical endmembers
for soil gas unaffected by geothermal gas, invoked to explain the observed soil gas concentrations. All endmembers have atmospheric concentration of He but CO2 concentrations
are different; atmospheric, 3 v%, and 10 v%.

wells and steam vents, respectively. The atmospheric composition
is depicted as a blue square in Fig. 6. Atmospheric CO2 in 2007 is
taken to be equal to 384 ppm V [46] and atmospheric He is equal to
5.2 ppm V [47]. The gas samples from the production wells are
taken to represent the deep geothermal gas in the Reykjanes
reservoir. The volume (and molar) ratio CO2/He in the deep
geothermal source is computed using the ratio between the
average CO2 and He concentrations in geothermal gas samples (i.e.,
23,700). The solid green curve in Fig. 6 represents the computed
mixing line between the atmospheric and deep geothermal
components.
Inspection of Fig. 6 shows that most soil gas samples plot close
to the mixing line between atmospheric and deep geothermal gas.
However a significant number of samples plot above the line and
several plot below it. Soil gas samples that plot above the mixing
line between atmospheric gas and deep geothermal gas
1302 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307

Fig. 7. CO2 and He in soil gas as a function of temperature at 45 cm depth. Horizontal lines show atmospheric concentrations of the gases þ10%.

Fig. 8. CO2/He ratio vs CO2 concentration in well and steam vent samples. Solid and dashed curves show predicted evolution of the steam composition during closed and open
system condensation.

composition call for another source of CO2 as processes that would
deplete the gas in He are most unlikely. The potential sources of
excess CO2 in this environment are the organic matter in the soil
and/or a “high-CO2” geothermal gas.
The CO2 concentrations in most soil gas samples are in excess of
atmospheric concentration; the CO2 concentrations of 4 samples
only are within 10% of the atmospheric value and 23 are within 50%
(see Section 4.1). The He concentration of a much larger group of
samples is near the atmospheric value; 200 out of the 305 soil gas
samples have He concentrations that lie within 10% of the atmo-
spheric value (see Section 4.1). Fig. 7 depicts CO2 and He concen-
trations as functions of temperature at 45 cm depth with blue and
red symbols, respectively. Horizontal lines indicate atmospheric
concentrations þ10%. Fig. 7 shows that CO2 concentrations in soil
gas in excess of atmospheric gas are found at all temperatures
whereas the elevated He concentrations are very rare at soil
temperatures below 50
C. Elevated CO2 concentrations in soil gas
with near atmospheric He concentrations at low temperatures
strongly indicate that decomposition of organic matter in the soil is
the source of the excess CO2 in the soil gas. This is consistent with
grass and moss vegetation cover in most of the study area, except
where the highest soil temperatures are observed. Unfortunately,
the carbon isotope composition of the CO2, that can be used to
discriminate between geothermal and biogenic sources of soil CO2
[15], was not determined in this study.
While organically derived CO2 is an obvious source of excess CO2
in soil gas particularly at low temperatures the presence of a “high-
CO2” geothermal gas component cannot be excluded. The presence
of this high-CO2 geothermal gas is inferred from the composition of
the steam vent samples. As noted in Section 4.2 the dry gas CO2
concentrations in four of the ten steam vent samples were higher
and the dry gas concentration of He and all other gases are lower
 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307 1303

Fig. 9. CO2 flux through soil as a function of CO2 concentration in the soil gas.

Fig. 10. Tracks of production wells and major feed points overlaid on a map of CO2 flux through soil (CO2 flux data from 2004 survey; [27].

than those of the steam from the production wells, while the
composition of the remaining six steam vent samples was very
similar to that of the well steam. These four samples were also
characterized by a higher total gas concentration in the steam
compared to that of the well steam and steam from other vents. The
locations of these steam vents on the periphery of the most intense
surface activity and the elevated concentrations of all gases (in mg/
kg steam) suggest that they have been affected by condensation. In
order to evaluate this possibility we performed simple model
calculations simulating the composition of the steam during
condensation at around 100
C. Two scenarios were considered
representing a closed and an open system, respectively.
Condensation calculations were carried out using the fitted
equations of Fern
andez-Prini et al. [23]; with liquid water density
calculated from the equations of Wagner and Pruss [57]. At 100
C,
the distribution constants for CO2 and He were calculated as 5.0 103
and 1.0 105, respectively. The distribution constant for CO2 is indeed
higher at all temperatures of interest, reflecting the fact that the
1304 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307
solubility of CO2 in liquid water is higher than that of He. The results
are shown in Fig. 8.
We simulated the condensation of steam sample 2007-0185
from well RN-14b, which is fairly typical for the Reykjanes area, at
100
C, using a liquid fraction step size of 0.001. The same behavior
was observed for both scenarios; the CO2/He mole ratio in the
vapor-phase gradually decreases with decreasing vapor fraction
from an initial value of 2.18  104, the only difference being that the
slope is steeper for the open system. At a vapor fraction of 0.05, the
CO2/He mole ratio is 1.95  104 for the open system and 2.17  104
for the closed system. Fig. 8 depicts the CO2/He molar ratio of the
steam vent samples (yellow) and the production well samples
(green) as a function of the total CO2 concentration. The evolution
of the CO2/He molar ratio as a function of CO2 concentration during
closed and open system condensation is shown by solid and dashed
curves, respectively. It is evident from Fig. 8 that neither conden-
sation scenario can explain the elevated CO2/He molar ratio in the
four anomalous steam vent samples. Partial closed system
condensation will not significantly change the CO2/He molar ratio
while open system condensation will result in a slightly lower CO2/
He ratio. This suggests that another process, in addition to
condensation must be invoked to increase the CO2/He ratio.
The process responsible for the elevated CO2/He ratio, as well as
the ratio of CO2 to any other gas considered, in the peripheral steam
vents cannot be unequivocally identified on the basis of the current
data set. However, it is clear that in addition to condensation that
increases the concentrations of all gases, CO2 must be added to the
vent steam from the four vents relative to the deep-reservoir steam.
This may be a result of the increased boiling in the system as a
result of pressure drop associated with the increased production
from the reservoir starting in May 2006. It is possible that the
increased boiling and steam flow from the reservoir may have
induced boiling in shallow CO2 rich aquifers or breakdown of calcite
in the cap rock. It seems unlikely that the excess CO2 in the steam
from vents is due to thermal breakdown of organic matter in the
surface soil although this possibility cannot be excluded in the
absence of carbon isotope data for the CO2 in the steam.
5.2. Low CO2/He soil gas samples: indication of CO2 retention in
calcite
One of the objectives of this study was to explore the possibility
of using the CO2/He ratio in the soil gas to quantify the amount of
CO2 that is retained in calcite while the steam ascends from the
geothermal reservoir. Calcite is common in the uppermost 1000 m
of the crust above the Reykjanes geothermal reservoir
[56,25,44,61]. Wiese et al. [61] quantified the amount of calcite in
249 cutting samples from 14 production wells in the Reykjanes
field. Based on these analyses they concluded that ~56 Mtons of CO2
were retained in the uppermost 700e1000 m, i.e. the bedrock be-
tween the geothermal reservoir and the surface (assuming an aerial
extent of 2 km2 for the reservoir). Using estimated age constraints
of the geothermal activity of 10,000 to 100,000 years they
concluded that the CO2 retention rate in the system was between
500 and 5000 t/year. When these values are compared to the
measured natural atmospheric CO2 emission in 2004 [27] of 5100 t/
year it is apparent that a significant fraction of the CO2 released
from the reservoir is retained in calcite in the bedrock above it. Such
a significant removal of CO2 from the geothermal gas would be
expected to result in a systematically lower CO2/He ratio in soil gas
relative to well gas. However, due to the apparent contribution of
decomposition of organic matter in the soil and the possible “high-
CO2” geothermal gas component discussed in Section 5.1 it is very
difficult to resolve systematic signs of CO2 depletion in the current
soil gas data set. It is, nevertheless, worth the effort to explore the
present soil gas results in this context.
Inspection of Fig. 6 reveals that most of the soil gas samples
collected in this study plot near the linear mixing trend between
deep geothermal gas (well samples) and atmospheric gas compo-
sition. The majority of the samples that deviate from this trend plot
above (or to the left) of the mixing line, most likely due to a CO2
contribution from organic matter in the soil as discussed in Section
5.1., above. However, there are several (~20) soil gas samples that
seem to be depleted in CO2 relative to He and plot significantly
below (or to the right) of the linear mixing line between these two
endmembers. These samples have the character expected to result
from significant CO2 retention in the bedrock. The data points to
which this applies plot along or above a mixing curve that repre-
sents a linear mixture between atmospheric gas and a geothermal
gas with a CO2/He ratio of ~5000 as opposed to ~25,000 observed in
the well samples. If this is interpreted as an indication of loss of CO2
then it corresponds to ~80% loss of the CO2 in the geothermal gas.
This indicates that effects of CO2 retention in the bedrock may be
detectable in the data set but due to the overprinting of the excess
CO2 that clearly affects the soil gas composition it is not possible to
make any quantitative estimation of CO2 retention from the current
data set. If the organic CO2 component in the soil gas could be
quantified using d13C analyses of the soil gas CO2 it might be
possible to “see through” the effect of excess organic CO2 and
quantify the CO2 retention by calcite.
5.3. Soil gas and temperature anomalies
The spatial distribution of soil temperature, He and CO2 con-
centrations in soil gas and CO2 flux, are shown in Figs. 2e5,
respectively. The comparison of the maps reveals that all the pa-
rameters delineate very similar surface anomalies. The hottest part
of the field is in the SE part of the study area, north of the road
between wells RN-1 and RN-3 (Fig. 2). This is also the location of the
most prominent CO2 flux (Fig. 5). Soil gas concentrations of CO2 and
in particular He are also elevated in this area although the soil gas
anomaly there is not as strong as the CO2 flux anomaly. This may in
part be due to the fact that soil gas samples could not be collected in
the most active part of this area (see absence of sampling points in
Figs. 3 and 4). There is also a distinct anomaly south of the east-
ernmost part of the lagoon. This anomaly is most prominent in the
soil gas concentrations of both gases although the CO2 flux anomaly
is quite strong in this area as well. Finally, there is a relatively weak
anomaly on the western border of the study area, by the lagoon.
This anomaly is most clearly seen in the soil temperature and to a
lesser extent in the CO2 and He concentrations. This anomaly is very
poorly defined by the CO2 flux data.
The anomalies defined by the CO2 and He concentrations in soil
gas are very similar in shape and intensity. This is not surprising
considering the correlation between the CO2 and He concentrations
in soil gas shown in Fig. 6. However, He soil gas anomalies are more
sharply defined than CO2 anomalies. This can for instance be seen
by comparing the soil gas concentrations around well RN-2. Around
that area and west of it the CO2 soil gas concentrations are some-
what elevated whereas the He concentrations in the soil gas are
close to atmospheric. The reason is that soil gas samples with at-
mospheric He and elevated CO2 are commonly encountered near
the margins of the thermal anomalies. This distribution is consis-
tent with the organic origin of some of the CO2 in the soil gas
postulated in Section 5.1. The good agreement between the
anomalies defined by the concentrations of these gases suggests
that in general He analyses may not add much to the information
gained by CO2 analyses of the soil gas. However, if weak CO2 soil gas
anomalies are detected their origin (in terms of geothermal vs.
biogenic) may be ambiguous then He analyses of the soil gas may
 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307
provide decisive complementary information.
Despite the relatively good overall agreement between the soil
gas anomalies (Figs. 3 and 4) and the CO2 flux anomalies (Fig. 5)
there are subtle differences between the results for these different
methods. As noted above, this difference is in part due to the fact
that soil gas samples could not be collected in the area with the
most intense surface activity where a high CO2 flux was detected.
But this is also the case in the second most prominent flux anomaly
south of the easternmost part of the lagoon. The most striking
discrepancy is about 100 m south of the lagoon where the highest
soil gas CO2 and He concentrations coincide with a very low CO2
flux. Similarly, on the western border of the study area modest CO2
and He soil gas anomalies are detected whereas the CO2 flux values
are very low.
Fig. 9, depicting the CO2 flux values for individual sampling
points as a function of CO2 concentration in the soil, illustrates a
surprisingly weak correlation between CO2 flux and the CO2 con-
centration in soil gas. It shows that even though low CO2 flux values
are likely to be observed at points where the CO2 concentration in
the soil is low there is nevertheless a spread of about two to three
orders of magnitude in the CO2 flux values that can be expected at
any given CO2 soil gas concentration. This is most likely due to the
heterogeneous permeability of the soil. In light of the weak corre-
lation between these parameters it is somewhat surprising to see
how well the anomalies defined by the two data sets agree
(cf. Figs. 4 and 5). This good agreement between the overall shapes
of the anomalies despite the weak correlation between the flux and
concentration values for individual data points (Fig. 9) suggests that
the heterogeneity of the soil properties is randomly distributed
over the study area. By using a sufficiently dense spacing of the
measurement points we are able “see through” the randomness
and thus obtain similar anomalies based on the different data sets.
It may be concluded that despite some minor differences be-
tween the anomalies delineated by soil temperature, CO2 flux and
soil gas concentrations it is quite striking that all parameters define
anomalies with more or less the same shape. This means that from
the perspective of geothermal exploration one or two of the pa-
rameters considered in this study give sufficient information. The
practical conclusion is to select the most rapid techniques, i.e. CO2
flux and soil temperature.
However, in those geothermal areas where no visible evidence
of gas emission is found at surface, the study of the spatial distri-
bution of soil helium offers an ideal geochemical tool to detect
faults controlling the emission of gases from deep origin due to the
geochemical characteristics of this noble gas [52]. Helium is highly
mobile, chemically inert, physically stable, non-biogenic, sparingly
soluble in water under ambient conditions, almost non-adsorbable,
and highly diffusive, with a diffusion coefficient about 10 times that
of CO2 [39]. Because of these properties, helium has been consid-
ered by geochemists as an almost ideal geochemical indicator. The
occurrence of deep source gases at shallow depths is controlled by
the presence of active faults/fractures, which are preferential
pathways for the discharge of gases toward the surface. Conse-
quently, soil helium surveys can help in identifying the most active
fracture systems in a specific area.
5.4. Gas flux anomalies and productive aquifers in wells
One of the objectives of mapping geothermal manifestations,
including soil temperature and diffuse gas flux is to delineate
structures, i.e. faults or fractures that permit steam to ascend from
the reservoir to the surface. The assumption is that these structures
extend more or less vertically to reservoir levels where they are
obviously feasible drilling targets. As there are not many producing
geothermal fields where diffuse soil gas flux has been mapped as
1305
extensively as in Reykjanes it is therefore of interest to compare the
CO2 flux anomalies in that field to the surface projections of the
productive aquifers or feed points in the production wells.
Seven out of eight directional production wells that have been
drilled in the Reykjanes field are directed towards the area where
soil gas flux has been measured. Fig. 10 shows the well traces for the
directional wells superimposed on a map showing the 2004 CO2
flux anomalies [27]. Feed points are indicated by brown symbols
and the symbol size indicates the relative magnitude of the feed
points (well traces and feed point locations are from the
ISOR
database).
The correlation between CO2 flux anomalies and feed points
depends to some degree on which flux data set is selected because
the details of the topography of the soil gas anomalies are slightly
different from year to year. This may be because of changes in the
system associated with the increased production from the
geothermal reservoir. Differences in the shape of the soil gas
anomalies between years may also be experimental artifacts due to
limited number of measurements (on the 25 by 25 m grid) relative
to the topography of the anomalies. Here the 2004 data set is
selected because it represents the CO2 anomalies as they were
before the commissioning of the Reykjanes power plant and the
westernmost anomalies (by the lagoon and the NS anomaly by well
RN-2) are better represented than in the data sets from later years.
It is evident from Fig. 10 that most of the productive aquifers in
the wells drilled under the study area are located immediately
below CO2 flux anomalies. The only major feed points that are
clearly located outside CO2 flux anomalies are the two major feed
points in well RN-26. Other feed points are located below flux
anomalies or their margins.
The most intense steam flow is from the SE part of the study
area, coinciding with the CO2 flux and soil temperature anomalies
north of the road between wells RN-1 and 3. It is worth pointing out
that the two wells that were drilled under that area, RN-22 and 26
do not have any important feed points there. This suggests that the
permeability anomaly that allows the steam to flow to the surface
in this area is either not vertical or spatially discontinuous at the
reservoir level.
The coincidence of gas flux, soil gas, and soil temperature
anomalies with major feed points in production wells in Reykjanes
suggests that detailed mapping of these parameters may be useful
for identifying feasible drilling targets in other geothermal systems.
Such data compliment conventional exploration methods particu-
larly surface mapping of faults and fractures, steam vents and other
surface manifestations. These methods are, of course, only appli-
cable where active surface manifestations are present, which is by
no means always the case. However, where such features are
accessible, detailed mapping of gas flux, soil gas, and soil temper-
ature may present an economically favorable opportunity to opti-
mize the success of drilling. For instance a CO2 flux and soil
temperature survey that covers the surface manifestations in the
Reykjanes field can be completed by two workers in less than one
week. The cost of such work is insignificant compared to the cost of
a geothermal production well.
5.5. Effect of production from the Reykjanes geothermal reservoir
on surface activity
Monitoring environmental changes or impacts resulting from
geothermal power production is among the responsibilities (be it
legal, contractual or moral) of geothermal power producers. These
impacts can result from power plant installations and civil works,
emissions of steam, gas, or brine or changes in surface activity
[2,41]. The methods used in this study are very useful for moni-
toring surface activity changes, in particular the CO2 soil flux
1306 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307
measurements that allow a quantitative estimate of the surface
activity changes between surveys.
Geothermal surface features are transient in nature; they are
invigorated from time to time by seismic events and decline
gradually in between (see e.g. Refs. [3,43]. Surface manifestations
are also sensitive to production from geothermal reservoirs.
Decline in discharge from hot springs as a result of production has
been described in several fields in New Zealand (see e.g. Refs.
[31,32,60] and the Philippines [11]. Experience from Rotorua, New
Zealand, has shown that such changes may be reversible to some
degree when production is decreased [53]. Steam vent activity and
diffuse heat flux through warm ground is, on the other hand,
known to increase in some cases as a result of production as the
resulting pressure drop leads to increased boiling in the reservoir
[33]. This has been the experience in Reykjanes where the surface
activity increased abruptly after the commissioning of the Rey-
kjanes power plant in 2006.
Surface activity in high temperature geothermal fields can be
broadly divided into point source emissions of steam or water and
diffuse activity. The diffuse activity is what we observe as
geothermal gas fluxes and elevated soil temperature and soil gas
concentrations. The manifestations of point source activity,
including steam vents, mud pits, and boiling springs are commonly
much more conspicuous than those of the diffuse activity. However
Fridriksson et al. [27]; showed that in 2004 more than 97% of the
CO2 emitted from the Reykjanes geothermal field was emitted
diffusely and similar results have been obtained in other
geothermal fields [22,59]. Based on these results it may be argued,
that invisible, diffuse gas emissions in geothermal fields represent
the most important parameter to quantitatively constrain changes
in surface activity in geothermal fields.
The pressure in the Reykjanes geothermal reservoir dropped
sharply when the 100 MWe power plant was commissioned in
2006. The pressure at 1500 m b.s.l. dropped from about 120 bar in
spring 2006 to about 95 bar in summer 2007 [28]. This pressure
drop caused a very noticeable increase in surface activity. Steam
vents were invigorated, soil temperature and CO2 flux increased
and soil temperature and soil gas anomalies expanded. The steam
flow from steam vents in the Reykjanes field in 2004 amounted to
0.8 kg/s [27] but in 2007 it had increased by almost an order of
magnitude, up to 6e7 kg/s [28]. The diffuse CO2 flow through soil
increased less drastically, i.e. by about 40%, from 13.5 t/day in 2004
to 18.5 t/day in 2007 (see Section 4.3).
Fridriksson et al. [27] used the measured diffuse CO2 flux
through soil to quantify the associated steam flow from the reser-
voir assuming that the concentration of uncondensed “primary”
steam from the reservoir, at atmospheric conditions was equal to
3.2 g/kg (see also Ref. [14]. The resulting value for steam flow from
the reservoir in 2004 was 49 kg/s. If the CO2 flux from the reservoir
in 2007 is used to compute the associated steam flow at that time
the resulting value is 67 kg/s. The total steam flow (from diffuse
emissions and steam vents) had thus increased from 50 kg/s in
2004 to 74.5 kg/s in 2007 (i.e. by about 50%).
6. Concluding remarks
Mapping soil gas concentrations, soil diffuse degassing and soil
temperature can provide very important data for geothermal de-
velopers. The experience in Reykjanes shows that these parameters
define very similar surface anomalies. At Reykjanes there is also a
strong correlation between productive aquifers in geothermal wells
and soil gas and soil temperature anomalies.
The similarity between the anomalies defined by soil gas con-
centrations, CO2 flux and soil temperature measurements suggests
that it may be sufficient to use only the determination of one or two
of these parameters for geothermal exploration. Considering that
the CO2 flux and soil temperature measurements take much less
time than soil gas sampling and analysis it seems that these
methods are more suitable for exploration purposes.
CO2 flux measurements have also the advantage to permit the
quantification of the magnitude of the soil degassing, thus allowing
quantifying the effects of production and/or of natural events, on to
the natural surface activity. Such data can be invaluable for quan-
titative evaluation of the magnitude of changes in surface activity
due to production or natural causes.
Because of the presence of biogenic CO2 in the soil at Reykjanes
and also because of the possible presence of anomalously high-CO2
steam in some steam vents at Reykjanes it was not possible to use
the data set collected in the current study to constrain CO2 reten-
tion in the bedrock above the Reykjanes geothermal reservoir. CO2
retention is expected to result in a lower CO2/He ratio in soil gas
relative to steam from the production wells. This trend was
observed for several soil gas samples but elevated CO2/He ratios
were found in the majority of the samples had, presumably because
of biogenic CO2.
Acknowledgments
Funding for this study was provided by Rannis e The Icelandic
Centre for Research, Reykjavík Energy, Hitaveita Sudurnesja (now
HS Orka hf), and the National Energy Authority. Bjarni Reyr
Kristja
Snæbjo

nsson, Sandra Osk €rnsdo

ttir, Niels Giroud, and Audur



Agla Oladottir assisted with field work. A review by Halldo
r


Armannsson greatly improved the manuscript.
References
[1] E.E. Archer, The determination of small amounts of sulphate by reduction to
hydrogen sulphide and titration with mercuric or cadmium salts with dithi-
zone as an indicator, Analyst 81 (1955) 181e182.


[2] H. Armannsson, H. Kristmannsdo
ttir, Geothermal environmental impact,
Geothermics 21 (1992) 869e880.




ttir, H. Torfason, M. Olafsson,
[3] H. Armannsson, H. Kristmannsdo Natural
changes in unexploited high-temperature geothermal areas in Iceland, in:
Proceedings of the World Geothermal Congress, Kyushu e Tohoku, Japan,
2000.
[4] S. Arno
rsson, Major element chemistry of the geothermal sea-water at Rey-
kjanes and Svartsengi, Iceland, Mineral. Mag. 42 (1978) 209e220.

rsson, Geothermal systems in Iceland: structure and conceptual
[5] S. Arno
models e I. High-temperature areas, Geothermics 24 (1995) 561e602.
[6] S. Arno
rsson, F. D'Amore, Monitoring of reservoir response to production, in:
S. Arno
rsson (Ed.), Isotopic and Chemical Techniques in Geothermal Explo-
ration, Development and Use, International Atomic Agency, Vienna, 2000, pp.
309e342.
[7]
rsson, J.O.
S. Arno € Bjarnason, N. Giroud, I. Gunnarsson, A. Stef
ansson, Sampling
and analysis of geothermal fluids, Geofluids 6 (2006) 203e216.
[8] S. Arno
rsson, A. Stef
ansson, J.O.€ Bjarnason, Fluidefluid interaction in
geothermal systems, in: A. Liebscher, C.A. Heinrich (Eds.), Reviews in Miner-
alogy and Geochemistry, vol. 65, 2007, pp. 259e312.
[9] S. Arno
rsson, G. Axelsson, K. Sæmundsson, Geothermal systems in Iceland,
€kull 58 (2008) 269e302.
Jo
[10] S. Bjo€ rnsson, B. Olafsd

ttir, J. To
o
masson, J. Jo
nsson, S. Arno
rsson,
S.G. Sigurmundsson, Reykjanes. Final Report on Investigations in the
Geothermal Area, National Energy Authority report, 1971, p. 173 (in
Icelandic).
[11] G.T. Bolan ~ os, E.V. Parilla, Response of Ban-Banati thermal area to development
of the Tongonan geothermal field, Philipp. Geotherm. 29 (2000) 499e508.
[12] C. Cardellini, G. Chiodini, F. Frondini, Application of stochastic simulations to
CO2 flux from soil: mapping and quantifying gas release, J. Geophys. Res. Solid
Earth 108 (2425) (2003), http://dx.doi.org/10.1029/2002JB002165.
[13] G. Chiodini, R. Cioni, M. Guidi, B. Raco, L. Marini, Soil CO2 flux measurements
in volcanic and geothermal areas, Appl. Geochem. 13 (1998) 543e552.
[14] G. Chiodini, D. Granieri, R. Avino, S. Caliro, A. Costa, C. Werner, Carbon dioxide
diffuse degassing and estimation of heat release from volcanic and hydro-
thermal systems, J. Geophys. Res. Solid Earth 110 (2006) B08204, http://
dx.doi.org/10.1029/2004JB003542.
[15] G. Chiodini, S. Caliro, C. Cardellini, R. Avino, D. Granieri, A. Schmidt, Carbon
isotopic composition of soil CO2 efflux, a powerful method to discriminate
different sources feeding soil CO2 degassing in volcanic-hydrothermal sys-
tems, Earth Planet. Sci. Lett. 274 (2008) 372e379, http://dx.doi.org/10.1016/
 T. Fridriksson et al. / Renewable Energy 86 (2016) 1295e1307
j.espl.2008.07.051.
[16] G. Ciotoli, M. Guerra, S. Lombardi, E. Vittori, Soil gas survey for tracing seis-
mogenic faults: a case study in the Fucino basin, Central Italy, J. Geophys. Res.
103 (1998) 23,781e23,794.
[17] A.E. Clifton, R.W. Schlische, Fracture populations on the Reykjanes Peninsula,
Iceland: comparison with experimental clay models of oblique rifting,
J. Geophys. Res. 108 (B2) (2003) 2074.
[18] A.E. Clifton, S.A. Kattenhorn, Structural architecture of a highly divergent plate
boundary segment, Tectonophysics 419 (2006) 27e40.
[19] C.V. Deutsch, A.G. Journel, GSLIB: Geostatistical Software Library and Users
Guide, Oxford University Press, New York, 1998.
[20] P. Einarsson, The seismic structure of Iceland, Jo €kull 58 (2008) 35e58.
[21] A.J. Ellis, W.A.J. Mahon, Chemistry and Geothermal Systems, Academic Press,
New York, San Francisco, London, 1977, p. 392.
[22] R. Favara, S. Giammanco, S. Inguaggiato, G. Pecoraino, Preliminary estimate of
CO2 output from Pantelleria Island volcano (Sicily, Italy): evidence of active
mantle degassing, Appl. Geochem. 16 (2001) 883e894.
[23] R. Ferna
ndez-Prini, J.L. Alvarez, A.H. Harvey, Henry's constants and vapour-
liquid distribution constants for gaseous solutes in H2O and D2O at high
temperatures, J. Phys. Chem. Ref. Data 32 (2003) 903e916.
[24] H. Franzson, Reykjanes High-temperature Geothermal System. Geological and
Geothermal Model, Iceland GeoSurvey report
ISOR-2004/012, 2004, p. 68 (in
Icelandic).
[25] H. Franzson, S. Tho
rdarson, G. Bjo € rnsson, S.Th. Gudlaugsson, B. Richter,

Fridleifsson, S. Tho
G.O.
rhallsson, Reykjanes high-temperature field, SW-
Iceland. Geology and hydrothermal alteration of well RN-10, in: Pro-
ceedings, 27th Workshop on Geothermal Reservoir Engineering Stanford
University, Stanford, California, 2002, p. 8.
[26] A.J.E. Freedman, D.K. Bird, S. Arno
rsson, Th Fridriksson, W.A. Elders,

Fridleifsson, Hydrothermal minerals record CO2 partial pressures in the
G.O.
Reykjanes geothermal system, Iceland, Am. J. Sci. 309 (2009) 788e833.
[27] Th Fridriksson, B.R. Kristja


nsson, H. Armannsson, E. Margre
tardo

ttir,


S. Olafsd
ttir, G. Chiodini, CO2 emissions and heat flow through soil, fumaroles
o G.I. Haraldsson, G.K. Halldo
and steam heated mud pools at the Reykjanes geothermal area, SW-Iceland,
Appl. Geochem. 21 (2006) 1551e1569.


[28] Th. Fridriksson, A.A. Olad
ttir, P. Jo
o
nsson, E.I. Eyjo

lfsdo

ttir, The response of the
Reykjanes geothermal system to 100 MWe power production: fluid chemistry
and surface activity, in: Proceedings of the World Geothermal Congress 2010,
Bali, Indonesia, 2010.
[29] W. Gander, Computermathematik, Birkh€ auser Verlag, Basel/Boston/Stuttgart,
1985, ISBN 3-7643-1688-8 (In German).
[30] W.F. Giggenbach, Chemical techniques in geothermal exploration, in:
F. D'Amore (Ed.), Application of Geochemisty in Geothermal Reservoir
Development, UNITAR/UNDP Centre on Small Energy Resources, Rome, 1991,
pp. 119e144.
[31] R.B. Glover, T.M. Hunt, Precursory changes to natural thermal features during
testing of the Wairakei and Broadlands-Ohaaki fields, in: Proceedings of the
18th New Zealand Geothermal Workshop, 1996, pp. 69e76.
[32] R.B. Glover, T.M. Hunt, C.M. Severne, Impacts of development on a natural
thermal feature and their mitigation e Ohaaki Pool, New Zealand, Geo-
thermics 29 (2000) 509e523.
[33] R.B. Glover, E.K. Mroczek, J.B. Finlayson, Fumarolic gas chemistry at Wairakei,
New Zealand, 1936-1998, Geothermics 30 (2001) 511e525.
[34] D. Granieri, G. Chiodini, W. Marzocchi, R. Avino, Continuous monitoring of CO2
soil diffuse degassing at Phlegraean Fields (Italy): Influence of environmental
and volcanic parameters, Earth Planet. Sci. Lett. 212 (2003) 167e179.
[35] B.Th. Gudmundsson, S. Arno
rsson, Geochemical monitoring of the Krafla and
Namafjall geothermal areas, N-Iceland, Geothermics 31 (2002) 195e234.
[36] V. Hardardo
ttir, K.L. Brown, Th Fridriksson, J.W. Hedenquist, M.D. Hannington,
S. Tho
rhallsson, Metals in deep liquid of the Reykjanes geothermal system,
southwest Iceland: implications for the composition of seafloor black smoker
fluids, Geology 37 (2009) 1103e1106.
[37] M.E. Hinkle, J.E. Kilburn, The Use of Vacutainer Tubes for Collection of Soil Gas
Samples for Helium Analysis, U.S. Geological Survey Open-File Report, 79-
1441, 1979.
[38] S.P. Jakobsson, J. Jo
nsson, F. Shido, Petrology of the Western Reykjanes
1307
Peninsula, Iceland, J. Petrol. 19 (1978) 669e705.
[39] A.C. Jenkins, G.A. Cook, Argon, Helium and the Rare Gases, History, Occurrence
and Properties, Interscience Publishers, London, 1961, p. 427.
[40] J.
nsson,
Jo Historic eruptions on the Reykjanes Peninsula,
Na
ttúrufraedingurinn 52 (1983) 127e139 (in Icelandic).
[41] H. Kristmannsdo


ttir, H. Armannsson, Environmental aspects of geothermal
energy utilization, Geothermics 32 (2003) 452e461.
[42] S.W. Lonker, H. Franzson, H. Kristmannsdo
ttir, Mineral-fluid interactions in
the Reykjanes and Svartsengi geothermal systems, Icel. Am. J. Sci. 293 (1993)
605e670.
[43] K.M. Mackenzie, P.R.L. Browne, M.P. Hochstein, Changes in thermal activity
of the northern Te Kopia geothermal field (NZ) during 1993-94, in: Pro-
ceedings of the 16th New Zealand Geothermal Workshop 1994, 1994, pp.
195e201.
[44] N. Marks, P. Schiffman, R.A. Zierenberg, H. Frannon, G.O.
Fridleifsson, Hy-
drothermal alteration in the Reykjanes geothermal system: insights from
Iceland deep drilling program well RN-17, J. Volcanol. Geotherm. Res. 189
(2010) 172e190.
[45] A. Mazot, D. Rouwet, Y. Taran, S. Iguaggiato, N. Varley, CO2 and He degassing
at El Chicho
n volcano, Chiapas, Mexico: gas flux, origin and relationship with
local and regional tectonics, Bull. Volcanol. (2011) 423e441, http://dx.doi.org/
10.107/s00445-010-0443-y.
[46] NOAA, 2011, ftp://ftp.cmdl.noaa.gov/ccg/co2/trends/co2_annmean_mlo.txt.
[47] B.M. Oliver, J.G. Bradley, H. Farrar, Helium concentrations in the Earths lower
atmosphere, Geochim. Cosmochim. Acta 48 (1984) 1759e1768.
[48] E.K. Padilla, Transport and Precipitation of Carbon and Sulphur in the Rey-
kjanes Geothermal System, Iceland, Master's thesis, Faculty of Earth Sciences,
University of Iceland, 2011, p. 58.
[49] E. Padro
n, D.L. Lo

pez, M.I. Magan ~ a, R. Marrero, N.M. Pe
rez, Diffuse degassing
and relation to structural flow path at Ahuachapa
n geothermal field, El Sal-
vador, Geotherm. Res. Counc. Trans. 27 (2003) 325e330.
[50] G. P
almason, G.V. Johnsen, H. Torfason, K. Sæmundsson, K. Ragnars,

rsson, Assessment of Geothermal Energy in Ice-
land, Orkustofnun OS-85076/JHD-10, 1985, p. 134 (in Icelandic).
[51] K.J. Parkinson, An improved method for measuring soil respiration in the field,
J. Appl. Ecol. 18 (1981) 221e228.
[52] L.A. Pogorski, G.S. Quirt, Helium emanometry in exploration for hydrocarbons,
Part I, in: B.M. Gottlieb (Ed.), Unconventional Methods in Exploration for
Petroleum and Natural Gas II, Southern Methodist University Press, Dallas,
1981, pp. 136e149.
[53] B.J. Scott, D.A. Gordon, A.D. Cody, Recovery of Rotorua geothermal field, New
Zealand: progress, issues and consequences, Geothermics 34 (2005) 161e185.
[54] M.A.
Sigurgeirsson, The younger Stampar eruption at Reykjanes,
Na
ttúrufraedingurinn 64 (1995) 211e230 (in Icelandic).
[55] M.A.
Sigurgeirsson, A chapter in the eruption history of Reykjanes: eruption
episode two thousand years ago, N
attúrufraedingurinn 72 (2004) 21e28 (in
Icelandic).
[56] J. To
masson, H. Kristmannsdo
ttir, High temperature alteration minerals and
thermal brines, Reykjanes, Iceland, Contrib. Mineral. Petrol. 36 (1972)
123e134.
[57] W. Wagner, A. Pruss, International equations for the saturation properties of
ordinary water substance e revised according to the international tempera-
ture scale of 1990 (vol 16, PG 893, 1987), J. Phys. Chem. Ref. Data 33 (1993)
783e787.
[58] V. Walia, S.J. Lin, C.C. Fu, T.F. Yang, W.L. Hong, K.L. Wen, C.H. Chen, Soilegas
monitoring: a tool for fault delineation studies along Hsinhua Fault (Tainan),
Southern Taiwan, Appl. Geochem. 25 (2010) 602e607.
[59] C. Werner, S.L. Brantley, K. Boomer, CO2 emissions related to the Yellowstone
volcanic system 2. Statistical sampling, total degassing, and transport mech-
anisms, J. Geophys. Res. 105 (2000) 10831e10846.
[60] P.A. White, T.M. Hunt, Simple modeling of the effects of exploitation on hot
springs, Geyser Valley, Wairakei, New Zealand, Geothermics 34 (2005)
187e207.


[61] F. Wiese, Th Fridriksson, H. Armannsson, CO2 Fixation by Calcite in High-
temperature Geothermal Systems in Iceland,
ISOR report
ISOR-2008/003,
2008, p. 68.