FLUIDS IN THE CONTINENTAL UPPER CRUST; BRINES FROM 4 KM DEPTH IN FRACTURED GNEISSES AT THE SITE OF THE GERMAN CONTINENTAL DEEP DRILLING PROJECT (KTB)

The 4 km deep pilot research drill hole of the German Continental Drilling Project has been used to sample fluids from the fracture pore space of granites and gneisses of the continental crust. The open hole and water entry points are located in the section from 3.85-4.0 km. The rocks are predominantly amphibolites and metasedimentary gneisses of the Bohemian massif. The Variscian metamorphic sequence is steeply dipping and complexly folded. The water has been extracted in the context of a pumping test of one year duration, during which water has been pumped with a constant rate of 29 l/min during the first 4 months and subsequently at 57 l/min. The temperature at depth is 120°C and 42°C at the wellhead. The fluid was sampled weekly during the test and the composition is essentially constant suggesting that the fluid reservoir is large and homogeneous. The test showed that water in the basement is apparently present in "unlimited" amounts. The hydraulic properties of the aquifer are similar to those of homogeneous porous media. The test also discovered a hydraulic boundary several 100 m away from the well at 4 km depth that is probably related to a major fault zone, the Franconian Lineament. The fluid is composed of 400 mmol/l CaCl2 and 280 mmol/l NaCl. XCa=0.6. TDS=63 g/l (seawater 35 g/l). The composition of the fluid is very different from other deep fluids in the Central European crystalline basement that normally has XCa<0.1 (e.g. Black Forest). We suggest that the main reason for the high XCa of the fluid is the presence of anorthite-rich plagioclase (An30 - An50) of the predominantly mafic rocks (amphibolites) of the basement. The Black Forest basement is mainly gneiss and granite with plagioclase < An20. This also implies that the principal reaction controlling the XCa of the fluid is plagioclase alteration. This is supported by low Cl/Br (ppm) of 73 (seawater 288), which strongly indicates a significant contribution from water-rock interaction. We propose that the original fluid was seawater that interacted at 120°C with andesine-labradorite of the basement amphibolite. The alteration increased the XCa of the water and excess Na formed new albite. The alteration reaction also produced abundant zeolite coatings on fractures of the crystalline rocks. Zeolite formation acted as an H2O-sink and as a result the TDS of the fluid has passively doubled "over the years".