Geothermics, Vol. 17, No. 4, pp. 619-633, 1988. 0375-6505/88 $3.00 + 0.

00
Printed in Great Britain. Pergamon Press plc
© 1988 CNR.

LOW ENTHALPY GEOTHERMAL FLUIDS FROM THE

PARIS SEDIMENTARY BASIN--1. CHARACTERISTICS
AND ORIGIN OF GASES

B. M A R T Y , * A. C R I A U D t and C. F O U I L L A C ?
*Laboratoire de Magmatologie et G~ochimie Inorganique et Expdrimentale, Universit( Paris 6, 4, Place
Jussieu, 75252 Paris Cedex 05, France and t lnstitut Mixte de Recherches G~othermiques, Bureau de
Recherches G~ologiques et Mini~res, 45060 OrlOans Cedex, France

(Received June 1987; accepted for publication February 1988)

Abstract Fifty-sevenwells tapping low enthalpy geothermal waters from the Dogger limestone reservoir
of the Paris Basin have been sampled and analysed for their gas composition. Methane/ethane ratios
indicate that hydrocarbons originate from both biogenesis and thermogenesis. Relatively high and
variable H2 concentrations are likely to result from fluid interaction with geothermal installations.
Regional trends among main species (N2, CO2, CH4) reflect large-scale heterogeneities which were
already revealed by the geochemistry of the fluids (Criaud et al., 1986). Northern sites are generally
nitrogen-rich and southern sites methane-rich, a characteristic which may be linked to the occurrence of
oil-bearing zones in the south. N2, Ar and He absolute contents show clear positive correlations. In
particular, NE-Artrends are best explained by the occurrence of a paleocomponent, likely to be an evolved
seawater. Helium model ages are consistent with the geological age of the host formation but are in
contradiction with hydrologic ages. Assuming an exotic flux of helium into the aquifer this discrepancy
may be overcome, but its computed rate apparently exceeds the current estimates for the continental
degassing flux.

INTRODUCTION
The Paris Basin provides a unique opportunity of extensive drilling for low enthalpy
geothermal resources. More than 60 geothermal doublets (production and injection wells)
tapping the Jurassic Dogger aquifer have been drilled since 1970 over a 200 km wide area. Most
of the plants are located in the central urban part around Paris (Fig. 1) and have provided heating
for over 500,000 people during the last few years (Roj as et al., 1987). Extensive hydrogeological
and chemical studies have been undertaken for different purposes: knowledge of the potential
resources, o p t i m u m development conditions, structure of the reservoir, problems caused by
injection after heat exchange or corrosional processes. As a result, the fluid geochemistry of the
Dogger aquifer is at present well documented (Iundt et al., 1983; Bastide, 1985; Criaud et al.,
1986). The geothermal fluids display highly variable salinity (6.4 up to 35 g 1-1) and chemical
composition. The t e m p e r a t u r e range is 47-85°C and the waters are mostly confined within three
limestone levels located between 1600 and 1850 m below the surface.
Gas geochemistry has been extensively used for reservoir engineering in high enthalpy
geothermal environments (e.g. D ' A m o r e and Panichi, 1980; Giggenbach, 1980; Giggenbach
et al., 1983; Arnorsson and Gunnlaugsson, 1985) but has been less often used in low enthalpy
geothermal reservoirs. Inert gas features have provided valuable information on the hydro-
dynamics of underground waters (e.g. Mazor, 1977; Andrews and Lee, 1979; Marine 1979;
H e a t o n , 1981). W e present here gas data from 57 wells, The aims of this study are (1) to provide
a complete set of gas data on a regional scale for an area already well documented for its geology
and water geochemistry and (2) to use these data for a better understanding of the origin,
chemical evolution and circulation of fluids in the Paris Basin.

619
620 h'. Marw, A. (riaud and ('. ~ m i l l a c

(" • Seine St_ Denis

NORTH ~ - Creil
k~' Beauvais
t~]A
© Val deMarne
SOUTH Le M6e
Fontainebleau
EAST [] Meaux
WEST A Ach~res-Cergy

f
BEAUVAIS

EVRY: V l
" | ~ ' ~ LE MEE-SUR-SEINE
../- ~,,,~'~Z~,~MELUN L'ALMONT

Fig. 1. Location of the sampled area. Sites having the same symbols show evidence of a specific evolution, according to
the chemistry of waters. Oil fields are exploited in the southeast border of this area. Stable isotope systematics indicate
a water recharge, which may occur at the eastern and southern borders of the Paris Basin (possibly the Morvan crystalline
outcrops).

G E O L O G I C A L SETTING
The sedimentary Paris Basin is a classical bowl shaped structure with an approximate radius
of 350 km and a maximum thickness of about 3000 m. The Paleozoic outcrops appear at the
eastern border (Vosges), to the south east (Morvan) and at the western border (Massif
Armoricain). The Dogger corresponds to a predominantly limestone lithostratigraphic
assemblage having a thickness of 100-300 m. The reservoir exploited for geothermal resources
cannot be described as a homogeneous unit, but rather corresponds to a complex stratified
structure, with a large number of thin producing levels, more or less continuous and showing
high but variable permeability (Rojas et al., 1987). The producing levels account for about
5-10% of the total thickness of the Dogger. Locally some good correlations between depth and
temperature (e.g. Val de Marne) exist and, in a specific depth interval (1600-1800 m) a high
standard deviation has been identified. A detailed analysis of temperature-depth relationships
is in progress (Menjoz et al., in preparation). Oil resources have been detected in the Dogger
and in the underlying Trias formations, east of the Paris urban concentration. In fact, most
geological prospections on the Dogger were initiated for oil exploration.
 Geothermal Fluids from the Paris Sedimentary Basin 621
GEOCHEMISTRY OF WATERS
The waters vary widely in composition and total salinity (TDS). The TDS variation does not
correlate with the bottom temperatures, reflecting complex mixing and circulation (Annual
Report, IMRG-BRGM, 1986). The occurrence of a seawater component modified by diagenetic
reactions was recognized from the unique CI : Br ratio for all the fluids (Bastide, 1985). However
the variations in boron or sulfate contents enable one to distinguish several groups among the
fluids, based on variations in the CI : B and C1 : SO4 ratios (Criaud et al., 1986). Sulfur studies
point to bacterial reduction of sulfates, as indicated by variable abundances of SO4 and the
isotopic composition of sulfur in sulfate. The 6180-6D relationship indicates that water recharge
occurs for some of the sites (Fouillac et al., 1986). The present geochemical sketch involves a
complex mixing of several components, including paleoseawater, young and dilute waters and
several evolution processes such as dissolution of evaporites, evaporation, diagenetic evolution,
bacterial reduction and water-rock interactions. These processes acted to various extents,
depending on their location in the basin, and different coherent areas were identified (Fig. 1).
(1) In Seine-St Denis county, north of Paris, the high sulfide content of the waters, the
isotopic composition of sulfate and the relatively low and variable sulfate contents provide
evidence of uniform bacterial reduction. This area is considered the most "confined", where the
fluids evolved in a more or less closed system (Fouillac et al., 1986).
(2) The southernmost group of Fontainebleau-Evry-Ris Orangis (referred to as Fontaine-
bleau from now on) and, to a lesser extent, Le M6e sur Seine-Melun l'Almont-Vaux le Penil
(Le M6e) group have the lowest TDS values, which are inversely correlated with temperature.
Aubertin and Ballin (1984) and Aubertin et al. (1987) report high hydrologic velocities (up to
~102 m yr -l at Fontainebleau) along a SE-NW direction. The study of 6180 and 6D (Fouillac
et al., 1986) points to a young recharge with low TDS in the southeastern part of the basin, thus
providing confirmation of the flow direction proposed by Aubertin and Ballin (1984), but
difficult to reconcile with the observed TDS-temperature relationship.
(3) The Meaux-Coulommiers (Meaux) area supplies the hottest and highest TDS fluids of the
basin. Isotope studies indicate that fluids are moderately evolved.
(4) Val de Marne County, south of Paris, is characterised by intermediate salinity and little
evidence of bacterial reduction. CI-B and CI-SO4 trends indicate that the fluids are the product
of an homogeneous mixing between Meaux-type fluids and Fontainebleau-type fluids.
(5) The western part of Paris and the fluids from the border of the basin cannot be classified
in the afore-mentioned groups, partly because few data are available for each individual area.

SAMPLING AND ANALYTICAL PROCEDURES

All gases were sampled at the production wellheads either at the end of the drilling test
(artesian flow) or later during production (under pumping). At the wellheads, pressures are in
the range 4-12 bar, thus corresponding to a single phase at depth for this type of fluid. A glass
separator operating at atmospheric pressure was connected to the wellheads and, after purging,
free gases were collected in glass vessels fitted with vacuum stopcocks. A flask vessel under
vacuum was half filled with the separated liquid phase, thus containing the gases dissolved or
present as microbubbles. During these operations flow rate was kept constant and the gas/liquid
ratio (GLR) was measured with a reproducibility of 5-10%.
Free and dissolved gases were analysed by gas chromatography with an accuracy of 3% in the
analytical section of the BRGM. The total gas content was computed from the analysis of the
free gas phase, the GLR and the determination of the dissolved gas content. The total CO2
content was computed from the gaseous CO2 content, pH and total alkalinity, which were
 b-J
Table 1, Gas content of the D o g g e r geothermal waters (mol I-~ ) cJ

Well Temp. TDS CO, N2 Ar He H2 CH 4 C2H 6 ('3H,

Site No. (°C) (gl-1) E-2 E-3 E-5 E-5 E-5 E4 E-5 E~5

North
Aulnay GAY1 71.6 24.3 1.08 2.35 3.9 4.33 12 6.05 2.97 8.02
Aulnay GAY3 70.7 25.0 0.91 2.72 2.33 4.38 4 6.31 4.5
Blanc-Mesnil GBMN1 67.3 26.5 0.86 1.35 1.09 2.06 0.06 4.8 2,1
Blanc-Mesnil GBMN 2 65.6 28.1 1.25 2.53 2.3 4.60 4.93 6.03 2.85 0.36
Clichy GCL1 70.8 19.4 0.89 1.40 1.4 2.50 0.00 19 19 146
Garges GGAR2 65.5 30.4 1.20 2.72 4.19 3.5 6.4 2.4
La Courneuve GLCN2 58.4 22.8 1.14 1.55 1.3 2.75 vat 5.05 2.2 2.4
La Courneuve GLCS1 56.6 21.8 0.97 1.55 1.4 2.80 var 6.85 3.3
Sevran GSAI 69.1 20.9 1.03 1.23 1.2 2.30 0.71 8.3 9.3 114
Tremblay GTRE1 73.9 26.1 0.80 1.77 1.3 2.00 0.04 10.2 8.4 49
Tremblay GTRE2 74.2 25.7 1.02 1.94 1.9 2.90 1.1 9 7
Creil GCRE4 58.0 28.0 0.86 1.77 1.7 2.67 3.92 6.3
Creil GCRE5 58.6 28.9 1.10 1.43 1.08 2.22 1.57 4.69 6.8 1.06
Creil GCRE6 56.4 28.2 1.00 1.09 1 2.61 7.2 5.45 7.34 1.24
Epernay GEPEI 57.3 8.8 0.68 0.88 1.46 1.65 0.06 2.48 6.42
Beauvais GBVSI 47.3 23.8 0.98 1.80 1.5 2.90 var 0,6 0.53
South
Bonneuil GBL1 79.3 22.1 0.87 1.46 1.47 2.76 3.16 29.1 19.5
Champigny GCHMI 78.2 25.3 1.05 1.05 0.6 1.85 3.36 16 11.3
Champigny GCHM2 73.9 25.6 0,93 1.40 0.9 2.36 2.76 17.7 13.3
Chelles GCHE1 70.2 22.7 0.88 1.35 1.51 3.00 0.07 30.4 21.4 211
Creteil GC R T 1 79.1 23.5 0.97 1.30 1.2 2.50 2.6 28 20 176
Crete il GCRT2 77.9 23.0 0.95 1.21 1.22 1.97 2.9 24.4 16.1 237
Epinay GESS1 62.5 14.1 0.77 1.31 1.88 2.84 0.05 29.5 19.4 135
Evry GEV1 71.8 10.4 0.95 1.85 1.55 3.50 0.02 24 15 128
Font ainebleau GFO 1 74.2 6.5 1.00 1.10 1.6 3.95 0.01 25 14.5 132
Ris-Orangis GRO 1 72.2 9.8 0.98 1.95 1.7 4.65 0.00 31.5 32.5
Le M6e GLMS1 73.0 12.9 0.90 I. 18 1.31 2.70 31.7 21.2 238
LeM6e GLMS2 1.02 1.64 1.56 3.50 29 17.9 389
East
Meaux GMX1 79.3 30.1 0.84 1.45 1.15 1.85 0.02 22.5 15 100
Meaux GMX3 76.5 31.8 0.89 1.30 1.08 1.40 0.6 21.5 13 91.5
Meaux GMX5 77.6 0.85 1.35 1.05 1.80 0.15 21.5 13.5 62
Meaux GMX6 78.0 33.0 1.02 1.38 1.92 2.05 9.9 21.1 11.3
Meaux GMX7 78.1 35.0 0.86 1.60 1.35 2.15 0.41 23.5 15.5 100
Coulom mie rs GCO 1 85.4 33.7 0.88 1.45 1.15 1.95 0.00 20.5 12.5 118
West
Cergy GCY2 51.0 13.2 1.18 1.80 2.82 3.45 0.07 5.63 5.82
 Geothermal Fluids from the Paris Sedimentary Basin 623

CO2

CH 4 A N2

Fig. 2. Relationship between gas concentrations of main species (symbols as in Fig. 1). Sites located at the northern part
of the study area are enriched in N2(end member B), those located in the south and east are methane-rich (end member
A). Both groups are more or less mixed with a CO2-rich pole (end member C). Beauvais sites, located at the northern
end, are almost devoid of methane (black triangles along the N2-CO2 axis).

CH4 biogenic ~ thermogenic 1OC2H e

Fig. 3. Relationship between carbon species. The same end members as in Fig. 2 encompass most of the data. The
biogenic-thermogenic transition along the methane--ethane axis corresponds to a C1/C1_2ratio of 0.960, according to the
classification given by Rice and Claypool (1981). The dotted line illustrates a possible trend for the Seine St Denis group.

d e t e r m i n e d on site. A noticeable o x y g e n c o n t e n t of 0.1 up to 5% is present in some of the

samples and is hardly compatible with the r e d u c e d character of the water, which contains
dissolved h y d r o g e n sulfide (up to 50 mg 1-1) and displays low r e d o x potential (Criaud et al.,
1988). O n l y a few ppb of dissolved 0 2 were detected on site using a colorimetric m e t h o d and the
real c o n c e n t r a t i o n must be m u c h lower. W e conclude that the o x y g e n originated outside the
aquifer and N2, A r contents were c o n s e q u e n t l y corrected for air c o n t a m i n a t i o n .
Because helium can diffuse t h r o u g h glass or stopcocks during vessel storage, helium data were
c h e c k e d by duplicating H e analyses using a different m e t h o d . H e samples were analysed the day
of sampling with a field H e mass s p e c t r o m e t e r ( M a r t y et al., 1987). G a s c h r o m a t o g r a p h y and
mass s p e c t r o m e t r y analyses were f o u n d to be similar within 5% for ten sites, that is, within the
precision levels for b o t h m e t h o d s .
624 B. Marl)', A. Criaud and C. Fouilhw
At the beginning of this survey, free gas sampling wets more systematic than total gas sampling.
resulting in a more comprehensivc data base for the former. For this reason, we use free gas data
(triangle diagrams, Figs 2 and 3) rather than total gas data when dealing with regional trends.
This data base is not reported in this paper but is available from the authors on request. The
absolute amounts of the gases are quantitatively used when evaluating the age and origin of the
waters and are therefore given in Table 1.

RESULTS AND DISCUSSION

Gas analyses display typical compositions of natural gases generally found in sedimentary
rocks, the major species being hydrocarbons (mainly methane), CO2 and N2 (e.g. Sokolov,
1974; Rice and Claypool, 1981). Hydrocarbons, H2S and part of N2 are thought to originate from
diagenetic evolution of fluids, biogenic and, possibly, thermogenic activities. Their contents (as
well as the CO2, H 2 concentrations) are related to the degree of evolution of the fluids at depth
and provide information on the thermodynamic conditions of the reservoir. Redox conditions
and corrosive processes are discussed in Criaud et al. (1988).

Reactive gases
Bastide (1985) demonstrated that thermodynamic equilibrium is achieved within the Dogger
reservoir. The concentration of components such as SIO2, Na, K or Ca and Mg is controlled by
a mineral assemblage including calcedony, calcite, kaolinite, dolomite and, surprisingly,
feldspar and albite. Gas composition in high enthalpy geothermal fluids is equally buffered by
mineral equilibria, which permits the derivation of gas geothermometers (Giggenbach, 1980;
D'Amore and Panichi, 1980; Arnorsson and Gunnlaugsson, 1985). Although the origin of the
gases in the present environment is not likely to emerge from the Fischer-Tropsch reaction, we
have tested gas geothermometry by computing temperatures with data from Table 1 and the
D'Amore and Panichi (1980) equation. However, computed temperatures are generally
50-80°C higher than downhole values, but temperatures as low as 37°C were also computed. The
discrepancies may result from several phenomena. On the one hand, important corrosion
processes in some sites are detected in the casing and tubing of the production wells and probably
account for the relatively high hydrogen concentrations (up to 0.9%) observed in Dogger gases.
These values are even higher than those encountered in typical high enthalpy fluids and cannot
result from temperature-dependent reactions. Moreover, the hydrogen content was found to
vary according to chemical treatment to inhibit corrosion (Ouzounian, personal communica-
tion). Introducing the measured H2 concentration in the D'Amore and Panichi formula, which
assumes very low hydrogen contents at medium to low temperatures, introduces substantial
errors in geothermometry computation. On the other hand, iron sulfide scaling may lower the
initial sulfide concentration of the water and gas, so that the gas geothermometer becomes less
reliable in the zones where these deposits and initial sulfide contents are important (e.g. Seine
St Denis). Finally, it is unlikely that the same mineral buffers as in high temperature geothermal
systems (generally in volcanic environments) could, in this particular case of a sedimentary
basin, control gas composition.
Total CO2 contents (Table 1) do not seem to be directly related to temperature, but equilibria
with calcite and aragonite may impose some control and more investigations are needed.
Applying the equations proposed by Arnorsson and Gunnlaugsson (1985) for the temperature
dependence of gas concentration leads to too high temperatures. Equilibria on oxydo-reduction
state are discussed elsewhere (Criaud et al., 1988).
 Geothermal Fluids from the Paris Sedimentary Basin 625
Regional trends
Total gas contents are shown in a CH4-COE-N 2 diagram (Fig. 2), revealing several groups that
are consistent with the chemical classification proposed by Criaud et al. (1986). The main groups
are the northern part of the Paris urban area (Seine St Denis, Creil) and the south-southeastern
area (Val de Marne-Le M6e-Fontainebleau). Other subgroup data are scarce and do not define
any specific trends, except for Meaux, which is intermediate, chemically and geographically,
between Seine St Denis and Val De Marne-Le M6e-Fontainebleau. The extreme point along
the N2-CO 2 axis corresponds to Beauvais, which is located about 100 km north of the central
area of the Dogger aquifer. Therefore, geographical variations show that the gases from the
Dogger aquifer are predominantly nitrogen-rich in the northern part (end member A in Fig. 2)
and methane-rich in south Paris (end member B). The two groups are more or less diluted by the
addition of CO2 (end member C).
In this sedimentary environment, biogenic and thermogenic production are likely to domi-
nate. Methane generation by micro-organisms can occur at depths that are equivalent to
temperatures between 0 and 75°C (e.g. Rice and Claypool, 1981), that is, for depths between the
surface and roughly 1500-2500 m. For higher temperatures, bacteria tend to degrade and
catagenesis becomes the major process responsible for methane production. Summarizing
worldwide data for biogenic gases, Rice and Claypool (1981) have outlined biogenic methane
characteristics according to the (~13C 1 and C1/C1-C2 ratios, the latter varying between 0.960 and
0.999. The Dogger productive layers are comprised in the biogenic-thermogenic transition
depth and gases have C1/C1-C 2 ratios ranging from 0.800 and 0.965. Therefore, part of the
hydrocarbons may have been formed by micro-organism production and part may have
originated from thermogenesis.
Figure 3 displays the CEH6 concentrations, together with the carbon dioxide and methane
contents. A comparison between Figs 2 and 3 shows that nitrogen-rich samples are also enriched
in ethane. In the case of the northern samples (Seine St Denis, black dots), gases tend to become
more "thermogenic" as the CO 2 concentrations decrease: starting from a CO2-rich, biogenic
pole, the data become progressively depleted in carbon dioxide and enriched in heavy
hydrocarbons. This trend probably reflects an evolution of the fluids upon thermal maturation
for the Seine St Denis area. Fluid maturation in these sites has been shown by the chemistry of
the waters (Criaud et al., 1986). In the southern sites the hydrocarbons are more abundant and
their ratios are rather constant (open symbols in Fig. 3). This contrast may be linked to the
proximity of oil fields in south Paris.

Nitrogen and rare gases

Argon, helium and nitrogen are supplied to the fluids according to three processes. Atmos-
pheric gases are introduced in underground aquifers as dissolved species in the recharge water
or in the water initially trapped in the formation (e.g. Mazor, 1977). Diagenesis leads to the
formation of N2 by reduction of NO3 and bacterial decomposition of nitrogen bound up in
organic compounds. Radioactive decay of uranium, thorium and 4°K produce 4He and 4°Ar
which are added to the fluids through the release of these nuclides from the formation rocks (in
situ production) or, as recently suggested (Heaton, 1984; Torgersen and Clarke, 1985), through
gas migration from deeper levels. Each process supplies a specific amount of inert gases, that can
be compared with the concentrations observed in the Dogger fluids.

(i) Nitrogen. Air-derived nitrogen in 0-25°C water amounts to 0.38--0.82 x 10 -3 mol 1-1
(salinity of 0 and 35 g 1-~; Weiss 1970). Heaton and Vogel (1981) have proposed that air could
be incorporated into an aquifer as microbubbles in the recharge water, substantially increasing
the amount of trapped atmospheric gases. Their best figure accounts for 10 ml air per litre of
626 B. MarO', A. ('riaud and C. k?millac
water, which extends the possible range for atmospheric nitrogen to 1.17 x l(I ) mol I ~. The
concentrations observed in the Dogger fluids are 0.88-2.72 x 10 -3 mol 1-~, thus showing that
part of the nitrogen does not derive from air. Its origin is likely to be linked with diagenetic
processes.

(ii) Helium. Because helium leaks from the atmosphere into space, its concentration in air
and, therefore, surface water is small. The concentrations in the Dogger fluids (1.4 A.5 X 10-~
mol 1-1) far exceed those of dissolved atmospheric helium (1.65-2.19 x 10-`) tool 1-1) and are
due to the radiogenic contribution of 4He. These concentrations are very high, but comparable
to those observed in sedimentary basins of comparable age (e.g. the Texas Pandhandle;
Wasserburg et al., 1963). He concentrations provide information on time-dependent processes
which are discussed in the last section.

(iii) Argon. By a similar comparison, observed concentrations (0.6-3.9 x 10-5 mol l - l ) are
similar or slightly higher than air-dissolved Ar contents (1.01-2.64 x 10-5 mol 1-1). Hence the
Ar contents can be considered to be of atmospheric origin, with a slight and variable contribution
of radiogenic Ar. To estimate the extent of this contribution, we first assume that the radiogenic
production occurred in situ, that is, within the host rock. The equation for the accumulation of
4°Ar in the fluid is

4°Ar/= O AAr ~ 4°K(u~t - 1)

where (0) is the density of the host rock, (q~) the porosity, (AAr) the release efficiency of 4"Ar
from the rock to the fluid (AAr may vary from 0 to 1), (2, 2e) the decay constants of (4°K ~ a°Ar),
(4°K) the number of atoms of potassium 40 per gram and (t) the duration of the contact between
the rock and the fluid. With appropriate constants the equation reduces to:

4°Arf = O AArKt x 1.738 x 10-15

where (K) is the potassium content of the rock (weight %) and (4°Arf) is expressed in mol 1-j.
Hence the production equation mainly depends on the contact duration between the fluid and
the rock, the potassium content of the rock and the release efficiency from the rock to the fluid.
As 4°Ar is produced following electron capture and subsequent gamma decay, the induced recoil
path length is likely to be short. Therefore the produced radiogenic isotopes should tend to have
sites similar to those of the parent isotopes and the release of 4°Ar into the fluid should mainly
proceed following solid, defect and grain boundary diffusions. This implies that the release
efficiency is not very high and variable according to the microscopic characteristics of the grains.
Secondly, the potassium content is very different when considering either limestones or shales.
For example, Zartman et al. (1961), summarising data from the literature, give average K
contents of 0.09 and 3.3% for these rocks respectively. Then, depending on which type of layer
contributes to the fluid (in the Dogger formation, productive 6oliths are interbedded with shale
sequences), the contribution efficiency for the release of radiogenic 4°Ar may vary greatly
according to the location. We can therefore only estimate the upper limit for this contribution,
by considering the age of the host rock (150 My) as the contact duration, a release efficiency
factor of one and the average potassium content of shales. These boundary conditions (together
with a density of 2.6 and a porosity of 0.15 for the 6oliths) lead to a radiogenic content of
1.5 x 10.5 tool 1-1, which is essentially similar to the concentration of dissolved atmospheric
argon and is in agreement with the observed concentrations. If argon is supplied from a source
outside the aquifer, as suggested in the case of helium (see below), its contribution could be
 Geothermal Fluids from the Paris Sedimentary Basin 627
derived from the inferred 4He exotic flux following migration (next subsection) and the
radiogenic 4He : 4°Ar ratio of 1-3 (e.g. Jochum et al., 1983). This gives an order of magnitude of
10-5-10 -6 mol l-t and is again comparable to the atmospheric contribution. We could therefore
conclude that argon mainly orginates from the atmosphere, with probable local contributions of
radiogenic 4°Ar. A more detailed analysis would require isotopic data that are not available at
present.

ORIGIN OF FLUIDS
As discussed above, Ar and N 2 contents in excess of air saturation values are due to the
radiogenic contribution of 4°Ar and to the biogenic production of,nitrogen. Lower contents may
result from previous outgassing events or a gas phase occurring at depth. Both possibilities are
unlikely in the case of the Dogger aquifer, considering downhole temperature-pressure
conditions and because production started recently. The dissolved gas concentrations depend on
the salinity and temperature of the water when it was in contact with air and their initial content
(before radiogenic and biogenic additions) are characteristic of the type of water which entered
the reservoir. As discussed earlier, the origin of the Dogger fluids may consist of the following
possibilities. (1) A paleoseawater component. The main facies of the fluid production layer are
of the 6olithic type and the probable temperature of a seawater of the same age as the
development of lagoons is around 25°C. Because seawater may have infiltrated during a colder
period, the possible lower range is taken to be 100C, the present temperature. (2) Assuming that
recharge occurs at the southeast border of the Paris Basin (Morvan Area, 300 km distance) as
suggested by stable isotope studies (Fouillac et al., 1986), hydrologic velocity estimates
(Auhertin et al., 1987) in the southern part of the study area give a recharge freshwater age of
103-104 yr. Considering the present average temperature and the fact that the earth surface
temperature was somewhat lower during the last 50,000 yr, the probable temperature range for
the recharge water is 0-20°C. Ar and N 2 contents are plotted in Fig. 4, together with the fresh
(recharge) water and seawater domains. Almost all data are consistent with the occurrence of an
initial warm seawater, but about half of them are not compatible with the dissolution of

Ar
1o-5mol/I o

N2 ~-3 mol/I
Fig. 4. Relationship between N 2 and Ar (absolute contents from Table 1, symbols as in Fig. 1). Bars represent
concentrations of these two gases in freshwater (FW) between 0 and 20°C and in seawater (SW) between 10 and 25°C
[solubilitydata from Weiss (1970)], plus the addition of 10 ml of air, in order to take into account the entrapment of air
microbubblesin the recharge water (Heaton and Vogel, 1981). Data are best explained by the occurrenceof moderately
warm seawater as an initial component. However, mixingwith a recent recharge cannot be ruled out as it will increase
'the Ar content by a factor depending on the mixingratio. The evolution of fluidsthrough diagenesisand agingincreases
both N2 and Ar contents.
628 B. Maro', A. Criaud and C. Fouillac
atmospheric gases in a cold freshwater as would be expected if the major component of thc
Dogger fluids was recently recharged. However, this does not mean that there was not a recent
recharge, but argues for an initial seawater component which may have been trapped during a
warmer period. Simultaneous increase of Ar and N 2in Fig. 4 reflects in this case the evolution of
the fluid by subsequent build-up of biogenic nitrogen and radiogenic argon. Any mixing with
recent freshwater would also increase the Ar content of the initial component by a famor
depending on the mixing ratio.

DATING PORE FLUIDS: DISCUSSION

Dating pore fluid is of particular importance in the case of the Dogger geothermal field which
is widely tapped for commercial purposes. Piezometric measurements and preliminary tracer
experiments have indicated that fluid circulation may be large (10-102 m yr -1) along a SE-NW
direction (Aubertin and Ballin, 1984; Aubertin et al., 1987). This direction is in agreement with
the general structure of the Paris basin and the inference of a present recharge at the eastern and
southern border of the basin. However, fast hydrologic velocities are hardly compatible with the
high salinity of the fluids (up to 35 g 1-1) which is best explained, from chemical evidence, by the
occurrence of a seawater component. Present-day seas (Atlantic Ocean and the North Sea) are
located on the NW side of the basin and cannot contribute to the reservoir according to the
general direction of fluid motion. Inert gas trends (Fig. 4) indicate a seawater trapping during an
epoch warmer than the present one. Finally, if recharge fluids form the main component, then
some explanation must be given of the path taken by these fluids.
Because radioactive decays are time dependent, He contents provide time constraints on the
possible evolution of fluids (e.g. Marine, 1979; Torgersen, 1980; Andrews, 1983; Heaton, 1984)
and can be used to estimate fluid ages. Radiogenic helium may be supplied to the system through
two main processes.
(1) In situ production. When the grain size of the pore rocks is below, say, 1 mm, 4He can
effectively escape due to its recoil during decay (100 micron length) and subsequent diffusion.
Therefore the helium content of the fluid reflects the U, Th content of the rock, the rock/fluid
interaction and the time during which the fluid remained in contact with the host rocks. Hence,
provided that other parameters can be safely estimated, a model age for the fluid can be derived
(Marine, 1979). Gamma logging tests have been achieved in several wells (Giot and Rojas,
1984). For the production layers, U and Th vary respectively between 1-3 ppm and 6-8 ppm.
These values do not differ much from average limestone contents (e.g. Marine, 1979). Three
main productive units have been recognized from flowmeter profiles and thermometry in wells
and the main productive unit, a white 6olithic formation, has a typical porosity of 15% (Rojas
et al., 1986). Using Marine (1979)'s equation, the in situ 4He production rate is computed to be
3.42 x 10-13 mol 1-j yr -1. On the other hand, the helium concentrations in the Dogger samples
range from 1.40 to 4.60 x 10-5 mol 1-I fluid, which gives production rates over 150 My (the
average age of the Dogger formation) of 0.93 to 3.07 x 10 13mol 1-1 yr -1. Low values can result
from mixing with recent waters or a more recent trapping of the seawater component, but higher
values are remarkably close to the computed production rate, despite the large uncertainties in
the age equation parameters. Model ages for Dogger fluids range from 10 My to 140 My with a
grouping around 70 My. The highest He contents are recorded in the Seine St Denis county, the
most "confined" zone. The in situ origin is illustrated in Fig. 5, which is a N2-He-Ar triangle
diagram adapted from Giggenbach et al. (1983). The radiogenic production of argon is taken
into account as the gases have an H e : A r ratio in the range of those predicted from typical
abundances of parent nuclides (Jochum et al., 1983). The data can be fairly understood as
resulting from aging of a groundwater (age in the order of 108 yr, area 1 in Fig. 5) and mixing with
a pure nitrogen pole, likely to be diagenetic in origin (area 2).
 Geothermal Fluids from the Paris Sedimentary Basin 629
(2) At variance with this interpretation are several indications that helium may originate
outside and deeper than the underground aquifers. Gas migrations at depth are common
(suspected for the Dogger in the case of hydrocarbons) and several authors report helium
concentrations in underground fluids in excess of those expected from other geochemical and
hydrological information. Heaton (1984), Torgersen and Clarke (1985) and Torgersen and Ivey
(1985) noted a systematic discrepancy between 4He ages and hydrologic ages in several aquifers.
They attributed these excesses to the release of helium from an exotic source and the
accumulation rate was computed to be 1.3 x 10 -11 mol 1-1 yr -1. This value is compatible with the
degassing rate of radiogenic isotopes from the solid earth as computed from global budget
considerations (e.g. Wasserburg et al., 1963; Ozima and Podosek, 1983). Other studies,
involving helium isotopes (Sano et al., 1986) and the relationship between heat and helium
(O'Nions and Oxburg, 1983) lead to similar degassing rates for 4He. From these similarities, the
source for exotic helium has been identified with the whole crust (Torgersen and Clarke, 1985).
In the case of the Dogger system, the He flux from an external source can be constrained
assuming that the hydrologic ages are correct. Taking a hydrologic age of 103-104 yr, a porosity
of 0.15 and an aquifer thickness of 20 m for the main productive unit (Rojas et al., 1987) leads to
a helium flux of 107-108 atoms (cm 2 s)-l. These fluxes are higher than the general estimates which
range from 2.8 to 7.2 x 106 atoms (cm 2 s) -1 (e.g. O'Nions and Oxburg, 1983; Sano et al., 1986)
but cannot be ruled out because flux calculation parameters are severely underconstrained.
We can see that the two possibilities discussed above, i.e. in situ production vs crustal He
accumulation, cannot be completely deciphered, but the first interpretation, which implies
negligible water velocities, has the advantage of being more precisely matched by He, Ar and N 2
data. Other information may be gained from a study of the stratigraphy of the whole formation.
Firstly, the Dogger productive units are underlain by massive shale deposits (50-100 m thickness
immediately below the lower sequence). Secondly, deeper aquifers are present in the Trias
formation; the He content of the Triassic fluid tapped at Melleray (100 km south of Paris) is
4.55 x 10 -5 mol 1-1, a value slightly higher but very comparable to the Dogger concentrations.
In order to obtain similar concentrations for helium in overlying aquifers, one has to assume a
steady state for the transfer of helium: the flux of He from the basement should enter the first
aquifer, then migrate from this table to a second aquifer, etc. This implies that He fluxes should

N2/100

He ~ Ar
radiogenic

Fig. 5. Inert gas relationship(data fromTable 1). N2, Ar and He are wellcorrelatedand can be understoodas resulting
from the mixingof (i) water dissolved-airgases, (ii) radiogenicheliumand argon (area 1), (iii) diageneticnitrogen(area
2). Dotted lines in area 1 represent water isochrons (ages givenin My).
630 B. Maro', A. ('riaud and C. Fouilh~c

be much larger than the flux computed in the case of accumulation. This view is not realistic as
the single accumulation flux already appears larger than possible fluxes computed by other
means. A possible alternative is that fluids in this sedimentary basin intercommunicate but again
this possibility is in contradiction with the "closed" nature of the aquifers evidenced from water
chemistry. In this sense, the Dogger gas characteristics pose more problems than solutions.

H E L I U M AND N I T R O G E N R E L A T I O N S H I P
Helium and nitrogen relative abundances display a good correlation, the N2/He ratio ranging
between 30 and 80 with a mean of 53 (_+ 13). The correlation is even better when considering
individual areas and gives an end member similar to the surface water (Fig. 6). These regularities
may be explained by two types of process that are not exclusive. Firstly, because organic matter,
the assumed source of some of the nitrogen, tends to concentrate the parent nuclides (U and Th)
of radiogenic helium, the observed correlation may indicate the evolution through time of more
or less organic-rich deposits. Secondly, nitrogen and helium show very comparable solubilities
(Weiss, 1970) and diffusivities in water (Arnold, 1930). Even if the gases migrate from deeper
levels, this transfer should not drastically fractionate N 2 and He because they are likely to occur
within an aqueous phase. Indeed, the most likely migration process for He is liquid diffusion in
interstitial fluids or displacement by groundwater motion because solid diffusion coefficients in
common mineral phases are too low. For example, at room temperature, the diffusion
coefficient for helium in minerals is in the order of 10 -l° ( c m 2 S)-I or lower (Barrer, 1951) as
opposed to 3 x 10 -5 (cm 2 s) -] (Arnold, 1930) in water, prohibiting displacements larger than a
few tens of meters during geological time. Gaseous diffusion may be limited to specific cases
(Newton and Round, 1961). In agreement with the preceding arguments, N2/He correlations
have been noted in different tectonic environments (e.g. Zartmann et al., 1961; Nagao et al.,

E
o

0
•I - 2-

initial water

| I

N~ , ~o"3 m o l / I

Fig. 6. N 2 and He absolute content relationship (data from Table 1). T h e initial water domain represents the
concentration of dissolved air N 2 with no radiogenic helium. Data from western and northern sites are fairly compatible
with the evolution of a trapped water component. Southern sites m a y have experienced a different evolution.
 Geothermal Fluids from the Paris Sedimentary Basin 631

1981). The constant N 2: He ratio over a large sample area can be interpreted as the result of the
degassing of a source wide enough to swamp local heterogeneities. The N2:He ratio in the
Dogger gases and the estimated 4He flux therefore are allowed to constrain the nitrogen flux in
the continental area of the Paris Basin. The fact that nitrogen and helium may be produced
within the reservoir or may migrate from deeper strata does not change this picture because,
providing a sufficient interval of time is available, gases will be released to the atmosphere in
both cases. For a 4He continental flux of 3 × 106 atoms (cm 2 s) -1 and an N 2 : He ratio of 53, the
continental degassing flux is therefore 1.6 × 108 atoms (cm 2 s) -1. Assuming that the Dogger case
is somewhat representative of the whole continental degassing, the corresponding N2 continental
flux would therefore be 4 x 107 mol yr -1 (30% continents). This value is negligible when
compared to the atmospheric abundance (2.76 × 1020mol) or even the sedimentary budget (very
variable with authors; Walker (1976) gives 102omol). This implies that the nitrogen stored in the
continental reservoir is not mobile. Sedimentary degassing may only marginally contribute to
the global cycle of atmospheric nitrogen.

CONCLUSIONS
(1) Gases from low enthalpy geothermal fluids in the sedimentary reservoir of the Parisian
Dogger appear to be a mixture of biogenic gases, thermogenic gases, radiogenic gases and
air-derived gases.
(2) Classical geothermometric functions derived from high temperature fluid studies (e.g.
D'Amore and Panichi, 1980; Giggenbach, 1980; and Arnorsson and Gunnlaugsson, 1985)
cannot be applied in this environment. Relatively high and variable H2 concentrations are best
explained by secondary processes linked with the economic recovery of geothermal fluids. This
inference is quantitatively discussed in Criaud et al. (1988).
(3) Gases display regional trends which are consistent with the chemical classification of
waters proposed by Criaud et al. (1986) for these areas. Sites located in north Paris are generally
rich in nitrogen and rare gases, while southern sites show greater concentrations of hydro-
carbons. Hydrocarbon migrations from oil-bearing zones are suspected for the latter.
(5) N2, Ar and He concentrations do not correlate with depth, salinity, temperature nor with
other volatile and dissolved species concentrations. However, these three gases display good
correlations over the whole study area. These correlations probably result from their inertness,
the similar physico-chemical behaviour of these gases during solution processes, their common
origin in the case of atmospheric N2 and Ar and from subsequent evolution of fluids through
diagenesis and aging.
(6) Inert gas trends are consistent with the occurrence of mixing between a paleocomponent,
possibly an evolved seawater, and a recent freshwater recharge, mainly in the southern part of
the study area.
(7) Helium dating of the fluids gives ages which are consistent with the geology of the host
formation, but not with the present hydrodynamic directions and velocities. This discrepancy
may be resolved by assuming an exotic He flux from deeper horizons. However, the estimated
flux necessary to match observed concentrations far exceeds the possible range of helium fluxes
from the solid earth. The structure of the Paris Basin in the central area is not very favourable to
the migration of helium and its accumulation in the Dogger reservoir unless the migrations of
fluids along permeable faults are considered. More investigations on fluid velocities, hydro-
geology and rare gas chemistry are therefore needed.

Acknowledgements--We are indebted to F. Iundt for initiating an early stage of this survey and to M. Brach for skillfully
assisting us on site. G. Ouzounian, C o m p a g n i e Frangaise de G6othermie, kindly allowed us to use unpublished data. W.
632 B. M a r l y , A . (?riaud a t t d ('. F o u i l l a c

F. Giggenbach greatly improved the presentation by suggesting the use of triangle diagrams and A. Jambon provided
useful comments. This work was supported by the Commission of the European Communities (contract n(~.
EN3G.0046.F(CD)) and also received support from the "Agence Fran~,aise Pour la Maitrise de l'Energie" (AFME) and
the BRGM. "Magmatologie et GOochimie lnorganique et Exp,Srimentale" (MAGIE) Laboratory is associated wilh
CNRS as UA736.

REFERENCES
Andrews, J. N. (1983) Dissolved radioelements and inert gases in geothermal investigations. Geothermics 12, 67-82.
Andrews, J. N. and Lee, D. J. (1979) Inert gases in groundwater from the Butter sandstone of England as indicators of
age and trends. J. Hydrol. 41,233-252.
Andrews, J. N., Goldbrunner, J. E., Darling, W. G., Hooker, P. J., Wilson, G. B., Youngman, M. J., Eichinger, L.,
Rauert, W. and Stichler, W. (1985) A radiochemical, hydrochemical and dissolved gas study of groundwaters in the
Molasse basin of Upper Austria. Earth Planet. Sci. Lett. 73,317-332.
Arnold, J. H. (1930) A kinetic theory of diffusion in liquid system. J. Am. Chem. Soc. 52, 3937-3955.
Arnorsson, S. and Gunnlaugsson, E. (1985) New gas geothermometers for geothermal exploration---calibration and
applications. Geochim. cosrnochim, acta 49, 1307-1325.
Aubertin, G. and Ballin, F. (1984) Analyse des 6coulements du Dogger du bassin de Paris. BRGM report
84 SGN 79 GTH-IRG, Orl6ans.
Aubertin, G., Benreditter, Y., Cordier, E., Doillon, F., Fabries, H., Gable, R., Gaillard, B., Ledoux, E. and de
Marsily, G. (1987) D0,termination experimentale de la vitesse d'6coulement de la nappe g~othermale du Dogger en
region Parisienne. SGFBuII. (in press).
Bastide, J. P. (1985) Etude g6ochimique de la nappe du Dogger du Bassin Parisien. Thesis, Universite de Paris 7, Paris.
Criaud, A., Fouillac, F. and Brach, M. (1986) Chemical evolution of the fluids from the Dogger aquifer. Proc.
Workshop Geochern., CEC Contractors, Anvers (in press).
Criaud, A., Fouillac, M. and Marty, B. (t988) Low enthalpy geothermal fluids from the Paris sedimentary basin--2.
Oxidation-reduction state. (Submitted to Geothermics.)
D'Amore, F. and Panichi, C. (1980) Evaluation of deep temperatures of hydrothermal systems by a new gas
geothermometer. Geochim. cosmochim, acta 44,549-556.
Fouillac, C., Fouillac, A. M., Criaud, A., lundt, F. and Rojas, J. (1986) Isotopic studies of oxygen, hydrogen and sulfur
in the Dogger from the Paris Basin. 5th Symp. Water-Rock Int. Trans., pp. 201-205, Reykjavik, Iceland.
Giggenbach, W. F. (1980) Geothermal gas equilibria. Geochim. cosmochim, acta 44, 2021-2032.
Giggenbach, W. F., Gonfiantini, R., Jangi, B. L. and Truesdell, A. H. (1983) Isotopic and chemical composition of
Parbati Valley geothermal discharges, North-West Himalaya, India. Geothermics 12,199-222.
Giot, D. and Rojas, J. (1984) Le Dogger dans les forages g6othermiques d'Aulnay Sous Bois et le Blanc Mesnil Nord.
BRGM Report 84 SGN 141 IRG, Orleans, p. 47.
Heaton, T. H. E. ( 1981 ) Dissolved gases: some applications to groundwater research. Trans. Geol. Soc. Afr. 84, 91-97.
Heaton, T. H. E. (1984) Rates and sources of 4He accumulation in groundwater. Hydrol. Sci. J. 29, 29-47.
Heaton, T. H. E. and Vogel, J. C. ( 1981) Excess air in groundwater. J. Hydrol. 50,201-208.
Iundt, F., Fouillac, C. and Ouzounian~ G. (1983) Etude g6ochimique des fluides du Dogger, premiers r~sultats. BRGM
Report 83 SGN 907 IRG, Orleans, p. 66.
Jochum, K. P., Hofmann, A. W., Ito, E., Seufer, H. M. and White, W. M. (1983) K, U and Th in mid-ocean ridge basalt
glasses and heat production, K/U, K/Rb in the mantle. Nature 306,431-436.
Marine, I. W. (1979) The use of naturally occurring He to estimate groundwater velocities for studies of geological
storage waste. Water Resource Res. 15, 1130-1136.
Marry. B., Brach, M., Criaud, A., Fouillac, C. and Vuataz, F. D. (1987) Analyse in situ de l'h~lium par spectrometrie
de masse 16g~re dans les fluides d'inter~t g~othermique. C. R. Acad. Sci.//305, 969-974.
Mazor, E. (1977) Geothermal tracing with atmospheric and radiogenic noble gases. Geothermics 5, 21-36.
Nagao, K., Takaoka, N. and Matsubayashi, O. (1981) Rare gas isotopic composition in natural gases of Japan. Earth
Plan. Sci. Lett. 53, 175-188.
Newton, R. and Round, G. F. (1961) The diffusion of helium through sedimentary rocks. Geochim. cosmochim, acta
22, 106-132.
O'Nions, R. K. and Oxburg, E. R. (1983) Heat and helium in the earth. Nature 306,429-431.
Ozima, M. and Podosek, F. A. (1983) Terrestrial Noble Gases. Cambridge University Press, Cambridge.
Rice, D. D. and Claypool, G. E. (1981) Generation, accumulation and resource potential of biogenic gases. A A P G
Bull, 65, 5-25.
Rojas, J., Menjoz, A., Martin, J. C., Criaud, A. and Fouillac, C. (1987) Development and exploitation of low enthalpy
geothermal systems: example of the Dogger in the Paris basin, France. 12th Workshop Geoth. Res. Eng. Stanford,
CA.
Sano, Y., Wakita, H. and Huang, C. W. (1986) Helium flux in a continental land area from 3He/4He ratio in northern
Taiwan. Nature 323, 55-57.
Sokolov, V. (1974) GOochimie des Gaz Naturels. Mir, Moscow (translated from Russian).
Torgersen, T. (1980) Controls on pore fluid concentration of 4He and ZZZRnand the calculation of 4He/222Rn ages. J.
geochem. Explor. 13, 57-75.
 G e o t h e r m a l Fluids f r o m the Paris S e d i m e n t a r y Basin 633

Torgersen, T. and Clarke, W. B. (1985) Helium accumulation in groundwater. I: an evaluation of sources and the
continental flux of crustal 4He in the Great Artesian Basin, Australia. Geochim. cosmochim, acta 49, 1211-1218.
Tor~gersen, T. and Ivey, G. N. (1985) Helium accumulation in groundwater. II: a model for the accumulation of crustal
"He degassing flux. Geochim. cosmochim, acta 49, 2445-2452,
Walker, J. C. G. (1977) Evolution of the Atmosphere. Macmillan, New York.
Wasserburg, G. J., Mazor, E. and Zartman, R. E. (1963) Isotopic and chemical composition of some natural gases. In
Earth Science and Meteoritics, pp. 219-240. North Holland, Amsterdam.
Weiss, R. F. (1970) The solubility of nitrogen, oxygen and argon in water and seawater. Deep Sea Res. 17,721-735.
Wilhelm, E., Battino, R. and Wilcock, R. J. (1977) Low pressure solubility of gases in liquid water. Chemical Rev. 77,
219-262.
Zartman, R. E., Wasserburg, G. J. and Reynolds, J. H. (1961) Helium, argon and carbon in some natural gases. J.
Geophys. Res. 66,277-306.