GRC Transactions, Vol. 34, 2010 Geochemical Investigations of Two Hydrothermal Systems in an Alpine External Crystalline Massif (Piedmont, Western Italy) Luca Guglielmetti 1,2, Romain Sonney 2, Giuseppe Mandrone 1, Eva Schill 2, and FrancoisDavid Vuataz 2 1 Department of Earth Sciences, University of Turin, Italy 2 Swiss Laboratory for Geothermics CREGE, University of Neuchâtel, Switzerland Keywords Thermal springs, geothermometers, water stable isotopes, geothermal reservoir, Argentera ABSTRACT The thermal springs of Bagni di Vinadio and Terme di Valdieri are located in the south of Piedmont, just 100km far away from Turin, in the Argentera Massif, the southernmost of the external Crystalline Massifs of the Alps. These waters have been geochemically studied to characterize the water path from the infiltration of cold meteoric waters to the discharge as thermal waters with temperatures up to 70 C. Water samples have been collected during two sampling campaigns, chemical and isotopic analysis were carried out to gather information about the chemical composition of cold and hot waters, the maximal equilibrium temperature reached at depth, the elevation of the recharge zone and the residence time. Thermal waters at Vinadio show a mixture of NaCl and NaSO 4 compositions and a wide variability in terms of Temperature and TDS while at Valdieri are very similar to each other showing a NaSO 4 chemical composition. Once cold meteoric waters infiltrate at more than 1700m a.s.l. through the faults they reach an average temperature of 120 C at Bagni di Vinadio and 100 C at Terme di Valdieri at depth. The circulation of the waters might take thousands of years to occur. Some hypothesis were developed to estimate the water rock interactions and the mixing processes in relation to natural fluid pathways inferred from the geological and structural settings of the study areas. Thermal waters interact with the rocks of the massif, in particular with granites and gneisses, that are involved in the tectonic activity of the massif. The chemical composition could suggest two different paths for the thermal waters discharging at the two thermal sites and also that mixing processes occur at Vinadio between a deep and hot NaCl end member and a NaSO 4 groundwater while at Valdieri the mixing process is reflected just in terms of diluition of NaSO 4 thermal with cold groundwaters. More information about the circulation pathways will be inferred from following geophysical and geological surveys that will help to better understand the very complex geological structures that control the circulation of these thermal waters and to define the most probable depth of the reservoir and eventually plan deeper investigations for the exploitation of these fluids. Introduction In Italy, only a few geothermal exploration projects were carried outside of Tuscany even though the northern part of Italy shows interesting geological settings where low enthalpy geothermal energy could be exploited. The southern part of Piedmont and in particular the Argentera Massif in the Italian Alps, presents a high concentration of hydrothermal circulations related to fault systems in the Western Alps. Therefore, favorable conditions for geothermal investigations of thermal waters are found at Bagni di Vinadio and Terme di Valdieri. These two sites present waters with different chemical compositions even though they are located in the same lithological setting (crystalline rocks of the Argentera Massif) and only 17 km apart. The thermal springs are concentrated within two small areas surrounded by high relief massifs. Figure 1. Geological sketch of the Argentera Massif. 689
The maximum discharge temperature reaches 70 C and 29 samples from thermal springs and boreholes, as well as cold waters, were collected and analyzed both chemically and isotopically ( 18 O, 2 H, 3 H) and data were interpreted to define the circulation models of these thermal waters. Geological Setting The two study areas are located in the Argentera Massif (AM) (Figure 1), the southernmost of the external crystalline massifs of the Alps. It is a popup structure cropping out owing both to its uplift and to the erosion of the Mesozoic sedimentary succession (Perello, 2001). The AM can be divided into two main complexes: the Malinvern Argentera Complex (MAC), mainly constituted by migmatitic gneisses, and the Tinée Complex (TC), mainly formed by anatectic gneisses. In particular the studied thermal springs are entirely located within the MAC, which rocks, mainly constituted of quartz, Kfeldspar, plagioclase, biotite, are highly milonitized and, as a result, the reservoir rock contains numerous microfractures (Michard, 1989). The structural setting of the Argentera Massif is the result of brittle reactivation of prealpine and earlyalpine ductile shear zones. A main NWSE shear zone, the Valletta Shear Zone (VSZ), crosscuts the massif dividing the two complexes and branches towards North in the Bersezio Fault (BF) that is located within the MAC. These two ductile shear zones, according to Baietto (2006) are interpreted as a unique complex fault zone named Bersezio Fault Zone (BFZ) that also comprehends brittle structures that often overprint the ductile zones. In the central part of the massif, the Valletta Shear Zone connects to the Fremamorta Shear Zone (FSZ), a EW oriented milonitic corridor with a reverse sense of shear. At Bagni di Vinadio thermal springs discharge at 1320 m a.s.l. through intensely fractured aplitic dykes located at the margins of the Bersezio Fault Zone, a 3 kmwide belt bounded to the NE by the Bersezio Fault and to SW by the Valletta Shear Zone. The Bersezio Fault Zone is a wide zone of fracturing and pervasive cataclasites composed by several faults trending 150 and consisting of fractured protholith rocks, fault breccias, cataclasites and gouges. The Terme di Valdieri springs discharge at 1425 m a.s.l. in one of the most elevated sector of the Argentera Massif. This area is pervasively cut off by NESW strikeslip faults with rightlateral slip components and by subsidiary left lateral ENE SWS striking faults. In particular the springs emerge through the damaged zone of the Lorusa Fault, a 7kmlong NWSE strike slip fault that cuts through migmatitic gneisses in association with the Cougne Fault that is more developed within granites. Water Chemistry Chemical analyses (Table 1) showed many differences between the thermal waters at Bagni di Vinadio and at Terme di Valdieri. At Bagni di Vinadio, thermal waters were sampled both in shallow wells and in thermal springs and show a NaCl composition, temperatures range between 26 C and 70 C and show a direct Figure 2. TDS vs T plot showing the direct relation between these two parameters at Bagni di Vinadio and the consistent conductivity at Valdieri despite the increase in temperature. meq/kg Figure. 3. a. Piper plot and b. Schoeller diagram for subthermal and thermal waters from Bagni di Vinadio and Terme di Valdieri. 690
Table 1. Selected chemical analyses of sampled waters. Sample Id Sampling Date Thermal Area UTM X UTM Y T C ph TDS Na + K + SVI1 16/7/09 Bagni di Vinadio 346734.7064 4906037.884 51.2 9.1 637 194.255 6.249 8.875 0.316 0.349 243.188 14.345 45.147 8.219 45.020 70.982 SVI1 25/11/09 Bagni di Vinadio 346734.7064 4906037.884 46.4 9.1 705 245.098 11.660 5.214 0.173 3.119 0.533 261.178 15.178 48.441 10.439 33.438 70.768 SVI2 16/7/09 Bagni di Vinadio 346746.6488 4906026.143 38.3 8.2 591 182.473 8.810 7.255 0.154 0.464 186.092 12.588 58.935 4.799 49.492 80.175 SVI2 25/11/09 Bagni di Vinadio 346746.6488 4906026.143 34.1 7.9 624 198.086 7.948 7.236 0.285 0.448 225.360 13.885 52.468 52.179 0.905 0.138 65.530 SVI3 16/7/09 Bagni di Vinadio 346739.8685 4906024.851 54.3 8.5 643 193.587 9.499 7.191 1.044 0.476 213.024 14.389 53.200 1.440 62.254 87.230 SVI3 25/11/09 Bagni di Vinadio 346739.8685 4906024.851 44.2 8.6 657 215.119 9.707 5.245 0.618 0.459 221.700 15.307 50.759 71.689 0.945 0.127 65.316 WVI1 16/7/09 Bagni di Vinadio 346720.1654 4905981.975 54.5 8.0 458 133.802 8.502 6.620 1.629 0.356 116.130 12.795 54.908 1.500 55.096 66.492 WVI1 25/11/09 Bagni di Vinadio 346720.1654 4905981.975 52.3 8.3 425 122.128 4.901 4.396 0.262 0.281 118.886 13.565 39.900 13.078 50.780 0.527 0.069 56.657 WVI2 16/7/09 Bagni di Vinadio 346738.4245 4906014.555 54.1 7.8 1175 373.878 24.035 27.428 0.156 1.054 606.215 11.403 10.616 0.000 45.815 74.830 WVI2 25/11/09 Bagni di Vinadio 346738.4245 4906014.555 51.0 8.8 1291 424.025 15.802 25.738 1.647 4.553 0.717 679.204 10.549 33.067 5.399 43.921 2.625 0.286 43.188 WVI3 16/7/09 Bagni di Vinadio 346770.2884 4905998.802 69.8 7.9 2348 711.645 34.149 63.104 0.116 2.071 1361.000 13.870 17.693 1.620 36.600 43.420 62.430 WVI3 25/11/09 Bagni di Vinadio 346770.2884 4905998.802 70.2 8.5 2724 902.224 47.814 75.446 0.908 4.621 2.158 1511.626 13.563 33.433 36.874 52.443 0.752 42.119 SVI4 16/7/09 Bagni di Vinadio 346816.1151 4905974.457 44.8 7.1 567 156.147 0.330 12.957 0.368 0.411 211.610 8.410 59.179 38.720 22.220 56.443 SVI4 25/11/09 Bagni di Vinadio 346816.1151 4905974.457 26.6 7.4 585 193.728 9.149 15.753 0.243 0.468 245.736 8.917 42.706 41.585 1.283 0.148 25.656 SVI5 16/7/09 Bagni di Vinadio 346871.6084 4906003.099 33.7 7.1 1272 415.001 18.778 31.185 0.141 1.183 667.988 8.991 33.433 34.028 61.361 SVI5 25/11/09 Bagni di Vinadio 346871.6084 4906003.099 38.8 8.7 1494 555.151 26.287 40.545 1.169 3.561 1.442 758.207 10.119 19.767 43.418 34.208 SVI6 16/7/09 Bagni di Vinadio 346802.1523 4905966.938 7.9 7.6 46 1.921 0.723 5.650 0.306 2.410 0.090 26.844 6.590 1.590 0.000 SVI6 25/11/09 Bagni di Vinadio 346802.1523 4905966.938 9.2 7.8 57 2.390 0.753 11.707 1.134 0.047 1.044 0.050 28.430 8.783 1.587 1.026 SVI7 25/11/09 Bagni di Vinadio 346651.6225 4905949.221 4.0 7.6 50 1.773 9.219 0.493 0.316 0.040 25.624 8.628 0.778 3.175 SVI8 25/11/09 Bagni di Vinadio 346692.1849 4905847.432 5.0 7.6 67 5.471 0.684 11.103 0.543 3.670 0.214 30.749 9.758 0.928 4.319 SVI9 25/11/09 Bagni di Vinadio 347090.1556 4905989.018 4.0 8.6 69 1.675 0.571 14.707 0.571 0.361 0.026 35.873 10.534 0.841 3.538 SVA1 17/7/09 Terme di Valdieri 361987.8369 4896466.827 34.3 9.4 351 82.782 3.164 5.011 0.433 0.187 27.800 11.577 29.284 12.779 74.453 103.265 SVA1 25/11/09 Terme di Valdieri 361987.8369 4896466.827 32.5 9.4 289 92.183 1.376 3.013 0.989 0.171 27.112 9.991 29.528 10.379 68.671 8.080 37.736 SVA3 17/7/09 Terme di Valdieri 361886.0318 4896525.24 44.5 9.5 317 77.806 9.541 4.717 0.024 0.163 30.597 11.186 31.847 11.219 62.443 76.968 SVA3 25/11/09 Terme di Valdieri 361886.0318 4896525.24 65.0 9.5 280 85.119 3.293 22.084 9.836 31.725 10.439 63.030 0.373 54.091 SVA4 17/7/09 Terme di Valdieri 361889.3697 4896507.716 33.0 8.9 176 27.691 2.100 13.680 0.809 0.058 15.363 5.366 44.171 3.540 31.392 31.642 SVA4 25/11/09 Terme di Valdieri 361889.3697 4896507.716 34.4 9.5 158 24.340 1.244 11.356 0.764 7.825 3.617 46.855 4.919 30.704 1.595 0.054 24.908 SVA5 17/7/09 Terme di Valdieri 361888.5353 4896511.054 54.1 9.5 309 74.285 3.138 7.834 0.764 0.155 24.685 11.374 29.040 12.779 69.244 75.471 SVA5 25/11/09 Terme di Valdieri 361888.5353 4896511.054 55.1 9.8 298 81.248 2.688 9.217 0.205 0.146 24.976 11.627 24.648 13.378 70.589 0.278 0.190 58.581 SVA6 17/7/09 Terme di Valdieri 361936.9344 4896546.102 53.0 9.7 341 88.044 3.457 4.236 0.013 0.199 29.481 12.492 36.361 16.498 67.451 82.313 SVA6 25/11/09 Terme di Valdieri 361936.9344 4896546.102 52.0 9.0 318 74.536 3.364 4.740 0.452 0.112 30.216 12.836 24.892 17.938 79.647 1.060 0.178 68.309 SVA7 17/7/09 Terme di Valdieri 361896.0455 4896512.723 46.4 9.7 296 73.467 3.124 4.941 0.047 0.155 23.557 10.188 33.189 14.218 58.209 74.616 SVA7 25/11/09 Terme di Valdieri 361896.0455 4896512.723 46.7 9.4 297 87.978 3.262 5.368 0.168 23.130 10.424 33.677 13.378 64.976 0.348 0.215 54.091 SVA8 17/7/09 Terme di Valdieri 361897.7144 4896478.51 56.9 9.6 311 76.896 2.836 4.943 0.023 0.166 24.995 11.444 36.361 14.938 62.224 76.540 SVA8 25/11/09 Terme di Valdieri 361897.7144 4896478.51 54.2 9.4 314 91.311 4.074 5.118 0.170 32.726 10.822 28.674 15.238 68.881 0.442 0.182 56.229 SVA9 17/7/09 Terme di Valdieri 361891.8731 4896469.331 57.2 9.7 298 71.984 2.293 4.036 0.007 0.140 24.801 11.471 29.406 15.478 61.669 76.540 SVA9 25/11/09 Terme di Valdieri 361891.8731 4896469.331 56.6 9.7 309 92.611 3.442 5.113 0.170 25.209 11.630 28.308 15.598 69.718 0.315 0.165 56.336 SVA10 17/7/09 Terme di Valdieri 361897.7144 4896506.047 58.1 9.7 196 46.520 1.480 4.810 0.140 0.100 4.032 1.967 34.409 14.578 11.516 76.754 SVA11 17/7/09 Terme di Valdieri 361902.7212 4896497.702 49.8 9.5 285 66.325 2.520 4.613 0.095 0.132 23.498 10.445 28.064 10.799 66.545 72.264 SVA11 25/11/09 Terme di Valdieri 361902.7212 4896497.702 44.7 9.7 291 87.367 3.530 5.880 0.148 0.166 23.691 10.377 27.210 13.738 66.459 0.348 0.132 51.526 SVA14 25/11/09 Terme di Valdieri 361923.0476 4896484.642 49.5 7.6 313 87.490 3.470 5.886 0.060 0.161 23.727 10.922 54.542 67.210 0.383 0.164 58.902 SVA16 25/11/09 Terme di Valdieri 362099.2117 4896667.155 20.5 7.3 147 35.437 1.402 5.597 0.273 0.066 9.616 4.497 30.505 2.040 28.678 1.140 0.085 27.794 SVA2 17/7/09 Terme di Valdieri 361955.2927 4896468.496 13.0 8.0 60 9.762 0.723 4.753 0.270 0.011 2.799 1.641 21.719 11.045 1.599 5.495 SVA2 25/11/09 Terme di Valdieri 361955.2927 4896468.496 11.2 8.5 47 7.392 0.655 1.919 0.506 2.884 0.214 12.934 10.325 5.192 0.034 4.661 SVA12 25/11/09 Terme di Valdieri 361933.3635 4896480.675 17.0 7.5 101 18.686 0.893 5.263 0.276 0.028 13.503 2.932 24.892 18.721 1.282 0.049 14.592 SVA13 25/11/09 Terme di Valdieri 361926.2217 4896470.359 17.1 7.6 93 17.898 0.834 5.335 0.309 0.025 5.043 2.765 23.427 18.017 1.302 0.039 17.767 SVA15 25/11/09 Terme di Valdieri 361883.3709 4896329.903 11.2 7.6 66 11.761 0.590 4.971 0.278 0.012 2.558 1.733 21.719 11.048 1.436 0.020 9.952 Ca 2+ Mg 2+ NH 4 + Li + Cl F HCO 3 CO 3 2 SO 4 2 NO 3 Br 2 SiO 2 relationship with the TDS that ranges between 400 and 2700 (Figure 2). Cold waters at Vinadio, sampled from cold springs and creeks at temperature between 4 C and 9 C, differ very much from hot waters, in fact they show a NaSO 4 composition and a TDS range between 4570. At Terme di Valdieri only hot springs are available for thermal water sampling and have Na SO 4 composition, temperature ranges between 32 C and 65 C but the TDS remains constant at about 300. Cold waters at Valdieri, sampled from cold springs and creeks, show a similar NaSO 4 composition even though the TDS lowers to 4590 and temperature ranges between 11 C and 17 C. Both the Vinadio and Valdieri waters have ph values of 79 or even higher and the alkalinity ranges between 0.2 to 1 meq/l at Vinadio and between 0.3 to 0.9 meq/l at Valdieri. The Piper diagram (Figure 3a) shows both the different chemical composition between thermal waters at the two study sites and the chemical differences between thermal and cold waters at Vinadio whereas at Valdieri these do not occur. A Schoeller diagram (Figure 3b) shows the general higher mineralization of the thermal waters, the high Cl content and the lower SO 4 concentration at Vinadio compared to that at Valdieri. The main cation is Na + for both waters at Vinadio and Valdieri but Cl is the major anion for Vinadio as SO 4 2 is for Valdieri. It is evident the differentiation between the thermal waters at Vinadio and at Valdieri: in fact Vinadio hot waters create a distinct group itself showing a higher concentration of chloride while those at Valdieri, richer in bicarbonate, are much closer to the cold water that have almost the same composition both at Vinadio and Valdieri. Mixing Processes Many thermal waters may mix with cold and less mineralized groundwater before being discharged in springs or wells. As a consequence sampled waters are usually a mix of 2 or 3 endmembers. At Bagni di Vinadio the thermal springs show a wide compositional variability and discharge a mixture of a NaCl endmember (probably related to the leaching of cataclastic hydrothermalized rocks) and a NaSO 4 endmember. This is evident comparing the thermal waters sampled from the deeper well (80m) WVI3 with the other thermal sources. The well discharges an almost typical 691
Figure 4. Plots showing the relation between TDS and the major ions and 18 O for both hot and cold waters at Bagni di Vinadio and Terme di Valdieri. NaCl water with the highest temperature (72 C), highest TDS (2800 ) but the lower SO 4 content (35) compared to its concentration (65) in the other thermal waters in the same area, demonstrating that the deep fluid could be rich in Cl and then mixes with NaSO 4 waters during its uplift (Figure 4). WaterRock Interaction Thermal waters circulate into a zone of intense fracturation and pervasive cataclasis, the Bersezio Fault Zone, constituted mainly by gneisses and micrograined granites containing quartz, kfeldspar, plagioclase, biotite, chlorite. Fluidmineral equilibria were computed with PHREEQ code choosing the most representative samples in terms of temperature and mineralization (WVI3 for Bagni di Vinadio and SVA3 for Terme di Valdieri). Saturation Indexes SI for sample WVI3 show that this water is in equilibrium with quartz, chalcedony, chlorite and fluorite, are undersaturated in respect to albite, dolomite, gypsum, halite, Kfeldspar, Kmica and oversaturated with respect to hematite, pyrite and sulfur. Water from sample SVA3 are in equilibrium with calcite, quartz, amorphous silica, chalcedony and quartz, undersaturated in 692 respect to albite, anhortite, dolomite, gypsum, halite, kfeldspar, kmica, kaolinite and oversaturated with respect to hematite, pyrite and sulfur. The origin of sodium could be inferred by the Na/Cl ratios. Both at Bagni di Vinadio and Terme di Valdieri it is positive (close to 2.53 for Vinadio and to 1.52 for Valdieri) indicating that the quantity of sodium is higher than chloride and then it excludes any contribution by hydrolysis of biotite. However sodium concentration could be related to rockfluid interactions in particular to the dissolution of the plagioclase contained in the migmatites. The origin of the Cl in the thermal waters at Bagni di Vinadio is not very clear among authors. Bortolami (1984) proposed that the high Cl content is related to mixing at depth of descending waters with connate brines injected inside the massif along tectonic lineaments, but as Perello (2001) and Baietto (2006) pointed out that all the recent cataclastic zones have a NWSE direction, transversal to the presumed direction of migration suggested by this author. Michard (1989) proposed that Cl could come from halite dissolution, but no halite is known to be present in the Argentera Massif even at depth so this hypothesis has to be rejected. Another source of Chloride was proposed by Arthaud and Dazy (1989) for thermal waters circulating along faults in the external crystalline massifs in the Western Alps. They suggest that the high concentration of Cl would come from the migration of old marine waters. Perello suggested in 2001 that Cl could derive from phyllosilicates and fluid inclusions contained in the crystalline rocks of the massif. Cl could replace the OH group in the phyllosilicates and it is very abundant in fluid inclusions since their major solute is NaCl (Roedder and Bodnar 1997). Since it is a very mobile ion it could concentrate in waters and hardly precipitate when enters the solution. At Terme di Valdieri thermal springs discharge a pure NaSO 4 endmember, the normal composition for waters circulating in the massif that probably originate through leaching of widespread granitic and migmatitic rocks containing minerals rich in SO 4 but also with relatively low Cl and Na + concentrations (Perello 2001). The only difference among these springs is the temperature that ranges between 32 C and 65 C in relation with the different degree of dilution occurred in the shallowest level during the thermal fluid up flow. Elevation of the Recharge Zone Isotopic analyses were carried out on selected water samples chosen from both thermal and cold waters. δ 2 H and δ 18 O content were evaluated to understand the origin of the waters in terms of typology and elevation of the infiltration. The contents of these two stable isotopes are given relative to the Standard Mean Oceanic Water (SMOW). The isotopic composition of thermal waters usually depends on physical processes, such as evaporation and condensation, that bring an enrichment in heavier oxygen isotopes. These processes occur in highenthalpy system but all the sampled waters plot close to World Meteoric Water Line and the Mediterranean Sea Meteoric Line (Pastorelli et al. 2001) indicating
that the thermal fluids are meteoric waters and in particular show the influence with Mediterranean precipitations (Figure 5). Both thermal springs and cold waters fall into the range between 10 and 13 ( vs. SMOW) for the isotope δ 18 O and within 80 and 100 ( vs. SMOW) for the δ 2 H isotope. Thermal waters do not exhibit positive δ 18 Oshift, typical of hightemperature geothermal systems with long water residence time at depths and where isotopic fractionation occurs because of boiling. Therefore the lack of positive δ 18 O shift indicates that the two studied geothermal systems are dynamic and fluiddominated systems. Water infiltration elevations can be inferred from these isotopic data: this relation varies according to the different geographical regions and, for the study areas those of Bortolami (1979) referring to the Maritime Alps were used. The calculation results show that the elevation quote is 15002000m a.s.l. for both Bagni di Vinadio and Terme di Valdieri with slightly higher elevations for Bagni di Vinadio. a b Figure 6. Saturation Index vs. Temperature Plots to estimate the reservoir temperature; a. Bagni di Vinadio, b. Terme di Valdieri. Figure 5. 18 O/ 2 H plot showing the relation between the stable isotopes of both thermal and cold waters. Temperature of the Reservoir Reservoir temperature is a very important data in particular in low enthalpy systems (<150 C) where is necessary to know very precisely the depth where might be possible to find exploitable temperature for energy production. Many chemical and isotopic reactions are temperaturedependant such as the fluidmineral equilibrium and the mineral assemblage in geothermal fluids, therefore they can be used as geothermometers to estimate reservoir temperature. Two main means are used to estimate the temperature of the reservoir: Saturation indexes of the main minerals Solute Geothermometers The variations of SI for samples WVI3 and SVA3 with increase of temperature were simulatied with the PHREEQ code to evaluate the reservoir temperature. SI vs. Temperature plots were created and showed an equilibrium temperature range between for 110 120 C for WVI3 and C and 95 110 C for SVA3 (Figure 6). Solute geothermometers of Silica (chalcedony), K 2 /Mg, Na/K/ Ca, Mg/Li, and Na/Li were calculated and compared to test their applicability (Table 2). They all show the same range of temperature even though some variations occur. Silica geothermometer of chalcedony was chosen because, for reservoir temperatures lower than 120 160 C as those present in the study area, chalcedony has a higher solubility than quartz (which works better for temperature range 150 225 C) and controls silica concentrations in the solution. Moreover just samples with ph lower than 8,8 were chosen for the chalcedony geothermometer because otherwise the temperature is overestimated. Na/K/Ca geothermometer is very effective and easy to apply but all the equations proposed by the several authors (Truesdell 1979, Arnorsson 1983, Fournier 1979, Giggenbach 1988) work for reservoir temperatures in the 180 350 C range but break down at temperature lower than 120 C. For this reason it tends to overestimate the temperature as well as the Na/K/Ca geothermometer if applied for lowtemperature reservoir (<120 C) and if not corrected as prescribed by Fournier and Potter (1979). Achievements of mineralsolution equilibrium can be assessed by means of the NaK Mg plot of Giggenbach (Figure 7). In geothermal systems at temperature lower than 150 C, as the 693
ones on Bagni di Vinadio and Terme di Valdieri, equilibrium is not attained and therefore the Na/K geothermometer could give wrong indications. All the samples plot below the full equilibrium curve, in the partially equilibrated and the immature waters fields. In this case K 2 /Mg geothermometer suits with the temperatures provided by the chalcedony geothermometer and in spite of some discrepancies these diagrams suggest temperatures of 110115 C for Vinadio and 100110 C for Valdieri. Moreover, hotter thermal waters are closer to the equilibrium curve while the colder thermal waters are closer to the points related to cold freshwaters at the Mg corner of the triangular plot, indicating a waterrock interactions at low temperature or more likely a mixing process with shallow, Mg 2+ rich, cold waters. Therefore it is possible to postulate that the chemical equilibrium conditions are reached at an average temperature of 120130 C at Vinadio and at 95105 C at Valdieri. Table2. Applied geothermometers to estimate the equilibrium temperature of the sampled thermal waters. Sample ID Chalcedony K²/Mg (Giggenbach) K²/Mg (Fournier) Mg/Li (Kharaka) Na/Li (low Cl) Na/K/ Ca 1 99 90 117 111 134 1 126 126 144 123 160 2 98 119 117 140 134 153 2 86 107 102 127 126 146 3 103 94 83 107 132 155 3 86 102 94 114 122 155 4 88 85 71 91 137 162 4 79 95 85 112 127 142 5 94 152 156 175 142 168 5 65 102 94 113 107 144 6 84 171 174 216 144 158 6 64 145 149 168 130 165 7 79 120 137 7 42 113 110 132 131 147 8 83 145 148 182 143 152 8 54 121 121 145 136 156 10 77 91 126 89 10 77 113 74 12 152 156 136 121 191 13 121 14 70 78 120 78 14 83 66 95 111 71 15 128 129 156 126 97 15 78 60 76 100 92 16 106 100 122 121 88 17 114 111 137 123 85 17 113 99 18 125 125 154 116 82 18 112 93 19 73 89 122 60 20 91 78 104 117 81 20 94 83 104 114 90 21 99 22 95 23 81 106 100 119 112 90 Average Vinadio 123 C Average Valdieri 104 C Figure 7. Giggenbach triangular plot showing how all the sampled water are below the fullequilibrium line. Groundwater Residence Time Residence time for geothermal fluids is usually evaluated by means of Tritium analyses. 3 H is the only radioactive isotope of Hydrogen and it has a halflife period of 12,43 years and is the most important isotopes for the study of groundwater. During the nuclear experiments in the early sixties the content of Tritium in the atmosphere increased up to 2000 TU. In thermal fluids with long residence time (waters infiltrated before the nuclear tests) its content is low and it can be a good indicator of the presence of mixing processes with younger waters infiltrated during or after the nuclear tests. Tritium analysis were carried out on selected samples comprehending the hottest samples (wells and hot spings) and also waters from cold springs. Figure 8. Tritium vs Chloride plot comparing the values from different authors. 694
The plot of Tritium vs. Chloride (Figure 8) shows that the samples with higher chloride content (~1500 ), therefore the hottest, have the lowest T.U content (0,6 T.U.). The highest content (6,3 T.U.) in the sampled waters, also compared to previous works (Michard 1989, Fancelli 1978, Regione Piemonte 1976), is on samples with lower Chloride (2,5 ) and therefore colder temperature (7,9 C).As observed the Tritium contents in thermal waters are always higher than 0.5 TU, the reference value for thermal waters circulating in crystalline massifs for long time, indicating that these waters are affected by mixing with younger groundwaters having a higher Tritium content. Conceptual Model On the basis of all the geochemical information gained from the analyses it is possible to suggest a model of circulation of the studied thermal waters. Local geothermal gradient values were estimated on the basis of the information of Botolami (1969) and Baietto (2006) suggesting an average gradient of 25 C/km. It is evident that the chemistry of the two end members is strictly related to the waterrock interaction and to the structural setting of the two studyareas. The two sites are just 17 km apart and the thermal waters come from meteoric waters infiltrating at about 1800 m a.s.l. through fractures and faults. At this point the two end members could begin to generate. On one hand the Bagni di Vinadio area is characterized by an intensely fracturated milonitic belt that could permit the waters to reach a slightly higher depth (3.54km), higher temperature (120130 C) than those of Valdieri. Therefore a stronger interaction with the host rocks by means of leaching, dissolution and interaction with fluid inclusions might occur, giving the waters the NaCl composition. Hot waters upflow to the surface loosing heat, in particular in the last superficial levels where they mix with NaSO 4 groundwaters and discharge at surface at the intersection between fault zones and the bottom of the valleys. On the other hand the thermal waters of Terme di Valdieri are much less mineralized than those of Vinadio and don t show any mixing process between end members with different chemical composition. The Valdieri area is less fracturated, therefore a less deep circulation system could involve the infiltrating waters. Descending waters could reach a 3km depth and a temperature of about 90 C and in these condition water rock interactions could occur less strongly. Na SO 4 upflowing hot waters mix with groundwaters of the same composition, they cool down and dilute discharging, as those of Vinadio, at the intersection of the Lorusa fault and the bottom of the valley. Conclusions At Bagni di Vinadio the thermal springs show a wide compositional variability and discharge a mixture of NaCl Vinadiotype end member and a NaSO 4 Valdieritype end member. At Terme di Valdieri thermal springs discharge a pure NaSO 4 endmember sometimes diluted by cold shallow waters. The compositional difference between the two end members cannot be related to a different mineralogical composition as the lithology at Vinadio and Valdieri is almost the same (gneiss and granite). In particular, there is not an evident and unique explanation for the high concentration of Cl at Bagni di Vinadio but, considering the tectonic and the geological setting, chloride could derive from phyllosilicates and fluid inclusions contained in the crystalline rocks of the massif. Bagni di Vinadio thermal waters reach an average equilibrium temperature of 120130 C and 95105 C those at Terme di Valdieri. Thermal waters from the both sites plot very close to the meteoric water line, indicating a pure meteoric origin. Infiltration takes long times and occurs thanks to the fault systems and water reach a depth of 34 km where get enriched in solutes in relation to the waterrock interaction conditions. Upflow occurs faster than infiltration as shown by the relatively high temperature of thermal springs that are located at the intersection between faults and the bottom of the valleys. Geochemical investigation gives useful information to plan further studies in these areas, in particular geophysical studies such as gravimetric and magnetotelluric surveys, to better indentify and characterize the fault systems and the reservoir as thermal waters in the studyarea could be exploited for direct use and maybe for power generation. References Arnorsson, S. (1983) Chemical equilibria in Icelandic geothermal systems Implications for chemical geothermometry investigations. Geothermics 12, 119128. Arnorsson, S (2000) Isotopic and chemical techniques in geothermal exploration, development and use. Vienna, Austria. International Atomic Energy Agency. Arthaud, F. & Dazy, J. (1989) Migration des saumures au font des chevauchementes de l arc alpin. CR Acad Sci Paris 309 (2) 14251430. Baietto, A. (2006) Faultrelated thermal circulations in the Argentera Massif (southwestern Alps). PhD Thesis, 3135. Bortolami, G. & Grasso, F. (1969). Osservazioni geologicoapplicative sul cunicolo d assaggio del traforo del Ciriegia e considerazioni sull intero tracciato Bortolami, G. et al. (1979) Isotope Hydrogeology of the Val Corsaglia, Maritime Alps, Piedmont, Italy. Isotope Hydrogeology, Vienna, Austria. International Atomic Energy Agency. Fancelli, R. & NUTI, S. (1978) Studio geochimico delle sorgenti termali del Massiccio Cristallino dell Argentera (Alpi Marittime). Boll. Soc. Geol. It., 97 115130. Fournier, R. O. (1991) Water geothermometers applied to geothermal energy. In Application of Geochemistry in Geothermal Reservoir Development. (F. D Amore, coordinator), UNITAR, 3769. Giggenbach, W. F. (1988) Geothermal solute equilibria. Derivation of NaKMg Ca geoindicators. Geochim. Cosmochim. Acta 52, 27492765. Michard, G. et al.. (1989) Influence of mobile ion concentrations on the chemical composition of geothermal waters in granitic areas, example of hot springs from Piemonte (Italy). Geothermics, 18 (5/6), 729741. Nicholson, K (1983) Geothermal Fluids. Chemistry and exploration techniques. Berlin, Germany: SpringerVerlag Pastorelli, S. et al. (2001) Chemistry, isotope values (δd, δ18o, δ34sso4) and temperatures of the water inflows in two Gotthard tunnels, Swiss Alps. Applied Geochemistry Volume 16, Issue 6, May 2001, Pages 633649. Perello P. et al. (2001) The thermal circuits of the Argentera Massif (western Alps, Italy): An example of lowenthalpy geothermal resources controlled by Neogene alpine tectonics. EclogaeGeologicaeHelvetiae, 94, 7594. Roedder, K. & Bodnar, R.J. (1997) Fluid inclusions studies on hydrothermal ore deposits. Geochemistry of ore deposits, 3 rd edition, Wiley, 657697. 695
696