PACROFI VI - Electronic Program

The Record of Multistage Fluid Flow in Recent Geothermal Quartz from Soultz-Sous-Forêts, France.

M.P. Smith1 , B.W.D Yardley1 & J.W. Valley2

  1. Department of Earth Sciences, University of Leeds, Leeds, LS2 9JT, U.K.
  2. Department of Geology and Geophysics, University of Wisconsin, 1215 W. Dayton St., Madison, Wisconsin 53706-1692.

Soultz-sous-Forêts lies in the Rhine Graben in eastern France, and is the site of the European Hot Dry Rock geothermal energy project. The site was chosen due to the high geothermal gradient (~10C/100m). Holes have reached depths of up to 3000m and pass through Triassic sandstones into Hercynian granite basement at depths of ~1400m (Aquillina et al., 1995). In hole EPS1, fluid at 150oC and with a salinity of ~10 wt. % NaCl eq. was produced from a large (~35cm thick) quartz vein at 2174m depth, in granite (Pauwels et al., 1993). Previous studies of quartz from this vein have shown it to contain 2 phase aqueous fluid inclusions with a range of salinities extending both above and below the modern fluid salinity but Th values that indicate fluid temperatures similar to the modern fluid (Dubois et al., 1996). The aim of the present study is to refine the quartz paragenesis within the vein and to attempt to correlate it with different fluid types in order to establish the chronology of fluids and to determine whether the vein is actively forming today. The sample used was a 100 micron thick quartz wafer from the vuggy central portion of the vein, supplied by J.J. Royer (Nancy).

Microthermometric measurements were made on fluid inclusions which were located accurately using both photographs and sketches of the sample. Fragments of wafer were then repolished and imaged using the cathodoluminescence (CL) mode of a Camscan series 4 SEM. Next the fragments were soaked for 48 hours in fluorosilicic acid in order to remove mica and carbonates. This also etched the surface of the sample so that the textural zones identified using CL could easily be distinguished using optical microscopy. The sample was broken into 61 distinct quartz domains, which were analysed for their oxygen isotope composition using laser fluorination at the University of Wisconsin.

Two main groupings of fluid inclusion salinity have been recognised in the vein. These are a population with 11 to 15 wt. % NaCl eq. (principally primary) and a mixed primary and secondary population with 4 to 8 wt. % NaCl eq., with a range of intermediate compositions. A series of texturally-distinct quartz types were observed and placed in chronological order using the SEM-CL images. The earliest generation is heavily fractured non-luminescent quartz (NLQ). This is overgrown by microcrystalline quartz which is cross-cut by fractures filled with a first generation of fine grained, euhedral, vuggy quartz (FGVQ1). All these generations are cross-cut by a central zone of coarse grained, euhedral, vuggy quartz (CGVQ) the crystals of which are surrounded by a coeval or slightly later second generation of fine grained, euhedral vuggy quartz (FGVQ2). Primary inclusions in the earlier quartz generations principally belong to the higher salinity population. The low salinity population is comparable to, or more dilute than, the modern fluid, and principally occurs in FGVQ2 and as secondary inclusion trails. In the earliest quartz (NLQ) only secondary inclusions are present, and these show the full range of fluid salinity. The majority of Th values range from 120-140oC with no correction for P, although a small number of inclusions homogenise at higher temperatures (up to 200oC). No measurements were made on fluid inclusions in the microcrystalline quartz.

The oxygen isotope data are shown in table 1. The NLQ shows the lightest values of d18O ranging from 10.05 to 13.22 permil. Microcrystalline quartz shows the heaviest values (15.02 to 17.83 permil), except for three analyses from the vein centre (here termed microcrystalline quartz 2) which show values comparable to the CGVQ and FGVQ2. CGVQ shows a very restricted range in d18O of 12.61 to 13.26 permil. The range for FGVQ overlaps this, but also extends to higher values (12.78 to 14.12 permil).

Quartz Typenmean d18O permil+/- 1sigmaRange
Non-luminescent quartz812.071.0710.05-13.22
N.L.Q. fracture fill212.340.8111.77-12.91
Microcrystalline quartz717.091.0015.02-17.83
Fine grained vuggy quartz615.811.2614.09-17.12
Coarse grained vuggy quartz2212.930.1512.61-13.26
Microcrystalline quartz313.150.0513.12-13.21
Fine grained vuggy quartz1313.300.4212.78-14.12

Table 1:- Mean oxygen isotope composition of different vein quartz generations from 2714m in the EPS1 borehole, Soultz-sous-Forêts, France.

The d18O values of the NLQ are closest to previous analyses of primary granitic quartz (8.7-9.8 permil; Yardley et al, 1995), suggesting that these fragments may result from brecciation and recementation of the vein wall rocks. The fluid inclusion population within this quartz thus represents the trapping of secondary inclusions as the quartz was repeatedly fractured throughout the formation of the vein.

Overall no correlation is seen between primary fluid inclusion salinity and quartz d18O suggesting that the isotopic composition of the fluid is controlled by equilibration with the host rock independant of original fluid source. The uniformity of Th suggests a relatively constant temperature hence thermal variations are unlikely to account for the wide range of d18O. Calculation of the composition of fluids in equilibrium with the different quartz generations at the modern temperature of 150oC using the fractionation factors of Zhang et al. (1989) shows that the coarse grained and second generation fine grained vuggy quartz may have formed from a fluid of similar isotopic composition to the modern fluid, despite their slightly higher salinities. These values may result from equilibration of the fluid with granite feldspar (Yardley et al, 1995). The heavier values seen in the first generation of fine grained vuggy quartz and in microcrystalline quartz imply a different fluid source. Precipitation of the microcrystalline and first generation FGVQ at 150oC would require a fluid with d18O from -1 to +3 permil. The heaviest analysed modern fluids from the area are those from the Bühl boreholes in Germany (d18O=-1.31 to -1.02) which tap fluids from the Triassic sandstones (Pauwells et al., 1993). The heavier quartz may therefore reflect periodic influx of fluids from the overlying sedimentary sequence which did not reach equilibrium with granite. Fluid inclusion characteristics from the FGVQ1 are not, however, distinct from the other quartz generations. The variation in fluid salinity has been interpreted as the result of mixing between low salinity surface fluids and brines derived from an evaporated seawater, probably formed by interaction of fluids with evaporite layers in the sedimentary cover (Pauwels et al., 1993; Dubois et al, 1995). The salinity data from this study supports such a hypothesis, and further suggests that the low salinity end member may have become more dominant with time.

Vein quartz from Soultz-sous-Forêt, France records a protracted history of fluid flow in its texture, its trapped fluid inclusions, and in its oxygen isotope composition. The salinity of fluid inclusions indicates that the fluid was sometimes more concentrated than at present, with the input of lower salinity, probably surface fluids probably increasing with time. The oxygen isotope composition of quartz indicates the fluid d18O was often controlled by equilibrium with feldspar in the host granite, although periodically heavier fluids, possibly derived from the overlying sediments, have entered the fracture system and precipitated quartz.