Tracer tests (artificial spikings) are indispensable in characterizing fluid-based geothermal reservoirs, in that they provide the only means for determining fluid residence times and fluid-rock contact-surface (i.e., heat exchange) areas; hydraulic and geophysical methods are largely insensitive w.r. to these parameters. 
  
Two single spikings and three complex spiking sequences (comprising single-well intra-layer push-pull, dual-scale push-pull, single-well inter-layer and inter-well flow-path tracings) were conducted in 4 km deep crystalline or sedimentary formations (candidate or actual geothermal systems) in Germany over the last five years (2003-2007). Main results were: 

-	at pilot KTB hole (intersecting permeable fault system in Mid-German crystalline basement): estimation of near-field and far-field fracture densities and of their increase/decrease during depletion/stimulation

-	at Urach-3 (HDR system in SW-German crystalline): superposition of tracer push-pull signals from at least two fractures with ambiguously-determined transport properties; interpretation remains unclear as yet

-	at Landau (hydrothermal system in mainly Buntsandstone layers, Upper Rhine Graben): improved estimation of least reservoir size (and thus of earliest thermal breakthrough) from lowered detection limits in tracer analytics

-	at Horstberg (water-induced frac in tight clay/sandstone layers in N-German sedimentary basin): estimation of inter-layer flow capture angle based on extrapolated tracer recovery; computing the flow-storage distribution of the induced hydrofrac based on a statistical time-moment analysis of the measured tracer breakthrough

-	from the ongoing tests at GroßSchönebeck (induced water- and gel-proppant fracs in volcanic and sandstone layers in N-German sedimentary basin): provisional estimation of frac volume and of fluid exchange with adjacent regions

  
In particular, an explanation is proposed for the relatively high tracer BTC tailing seen in the Horstberg flow-path tracing (from which signal background contributions could be excluded with certainty), which appeared as irritating at first from the point of view of local geothermal modeling.

  
Time-rate design options of continued fluid sampling and of new spiking operations are proposed and analysed for the KTB, Horstberg/Hannover, GroßSchönebeck, Landau and Bruchsal sites.

  
ACKNOWLEDGEMENTS: to the German Research Foundation (DFG) and the German Ministry for Environment (BMU) for financing most of the tracer experiments (under grant nos. Sa-501/16/1-4, Sa-501/21/1 and BMU-FKZ0327579, respectively), to BGR and GGA Hannover, to the GeoForschungszentrum Potsdam, to the BESTEC-for-nature and geoX Societies for financial support and fruitful cooperation.

Vilnius, Lithuania