Una nueva interpretación cuantitativa de curvas de avance con mesetasy largos limbos de recesión a partir de ensayos de trazadoren el acuífero cárstico artesiano de Stuttgart, Alemania

Abstract

In 1998 and 1999, two multi-tracer experiments were conducted in the artesian karst aquifer of the mineral springs of Stuttgart, Germany. The breakthrough curves (BTCs) monitored at the springs showed very long tails or developed plateau-like concentration levels for more than 200 days. Initially, this observation was qualitatively explained by exchange between cavities with stagnant water and the active conduits. Since then, a new analytical solution for tracer transport in karst aquifers has become available, the "two-region non-equilibrium model" (2RNE), which assumes the presence of mobile and immobile fluid regions, and mass transfer between these two regions. The experiments were thus revisited, and it was possible to provide a more quantitative explanation of the observed behaviour. The new model simulated all BTCs very well, thus confirming the earlier qualitative explanation. The prolonged BTCs can be attributed to intermediate storage in cavities containing quasi-immobile groundwater, and slow release into active fractures and conduits. The results also demonstrate that karst aquifers are not always fast-flushing systems, but contaminants can sometimes remain in immobile fluid regions for long periods.Résumé En 1998 et 1999, deux tests avec des traçeurs multiples ont été effectués dans l'aquifère karstique des sources minérales de Stuttgart, Allemagne. Les courbes de restitution (BTCs) observées aux sources ont montré des niveaux de concentration à queue très étirée et à profil en plateau développées pendant plus de 200 jours. Initialement, cette observation a été expliquée qualitativement par un échange entre des cavités avec de l'eau stagnante et des conduits actifs. Depuis lors, une nouvelle solution analytique pour le transfert de traçeur dans un aquifère karstique est devenue disponible, le modèle hors équilibre à deux régions (2RNE), qui suppose la présence de régions à fluides mobiles et immobiles, et un transfert de masse entre ces deux régions. Les tests ont été réexaminés, et il a été possible de fournir une explication plus quantitative du comportement observé. Le nouveau modèle a très bien simulé tous les BTCs, confirmant ainsi l'explication qualitative antérieure. Les BTCs prolongées peuvent être attribuées à un stockage intermédiaire dans des cavités contenant de l'eau souterraine quasi immobile, et une restitution lente dans des fractures et des conduits actifs. Les résultats démontrent aussi que les aquifères karstiques ne sont pas toujours des systèmes à chasse rapide, mais les polluants peuvent parfois demeurer dans des zones de fluide immobile pendant de longues périodes.Resumen En 1998 y 1999 se llevaron a cabo dos ensayos de trazadores múltiples en el acuífero cárstico artesiano que alimenta manantiales minerales de Stuttgart, Alemania. Las curvas de salida (CS) registradas en los manantiales mostraron colas muy largas o concentraciones en forma de meseta durante más de 200 días. Originalmente, esta observación fue cualitativamente interpretada por el intercambio entre cav...

Inversion for the Input History of a Dye Tracing Experiment

“…The Toride et al model dealt with the case in which there is a strong solute interaction between mobile-and immobile-flow regions or aquifer kinetic reactions, while our model focuses on the case in which aquifer kinetic reactions and solute interaction between mobile-and immobile-flow regions are negligible. Our solution to the ADM may be regarded as more robust, because it requires fewer parameters, which is similar to the findings of Gö ppert and Goldscheider (2008). In this sense, our model may be more suitable to the relatively rapid transport conditions that typically occur in solution conduits, because when solutes are rapidly flushed and entrained through the solution conduit, only a small portion of the solute will likely be sequestered in an immobile-flow region (Li et al, 2008) and probably only for a very short time.…”

“…Currently, the ADM and the two-region non-equilibrium model (2RNE) are two of the most popular models used for simulating solute transport in solution conduits (Toride et al, 1993;Field and Pinsky, 2000;Birk et al, 2005;Gö ppert and Goldscheider, 2008;Goldscheider, 2008). The 2RNE is advantageous in that excessive BTC skewness can be well simulated, but has the disadvantage of requiring additional parameters that lead to mathematical complications and possible errors in parameterization if local minima (as opposed to a global minimum) create a condition of non-uniqueness during inverse analysis (see, for example, Moré and Wright, 1993).…”

Drainage and siphoning of a karstic spring: a case study

“…However, we cannot discount that there may be other siphons operating upstream of the cave, as stated above. At least in this segment of the aquifer, it does not make sense to speak of groundwater velocity, because this velocity depends on when the recharge event occurs, a characteristic observed in other karstic aquifers (Field and Pinsky, 2000;Goldscheider N., 2005;Goldscheider N., 2008).…”

Scale Analysis of the Significance of Dispersion in Mixing-Transport in Cond

“…Equations including advection and dispersion were developed to account for retention in immobile-fluid regions (2RNE) and describe transport in a single conduit (Field and Pinsky, 2000;Goldscheider, 2008). The idea is use of the Green's function for the boundary value problem to model solute transport in a conduit.…”