Transport of three herbicides in ground water at Twin Lake test site, Chalk River, Ontario, Canada.

A field tracer test performed under natural flow conditions at the Twin Lake test site, Chalk River Laboratories of the Atomic Energy of Canada Ltd. in Chalk River, Ontario, Canada, using tritium and three herbicides (Chlortoluron, Terbuthylazine, and Pendimethalin) was interpreted using the dispersion equation with a combined reaction model. The reaction model couples an instantaneous equilibrium reaction governed by a linear adsorption isotherm with a reversible or irreversible kinetic reaction of the first order, and decay. An improved interpretation method consists of a simultaneous fitting of theoretical concentration and mass-recovery curves to the experimental data, which leads to a more reliable determining of reaction models and improves the accuracy of fitting. Tritium served as the reference tracer to determine the flow velocity, dispersivity, and the recovery of the herbicides. Chlortoluron was slightly delayed by equilibrium exchange with strongly reduced concentration due to an irreversible kinetic reaction and/or decay. Terbuthilazine was slightly delayed by equilibrium exchange, with strongly reduced concentration due to a reversible kinetic reaction with some influence of decay. A strong equilibrium reaction and a strong reversible kinetic reaction without degradation governed the transport of Pendimethalin, reducing considerably its concentration. The results obtained show that simulations based only on Kd and decay constant, especially if these parameters are found in the laboratory, may considerably differ from those performed with reaction parameters determined in properly performed field tests. The dominant reaction types, and the values of parameters found in the study, supply useful information on the transport of the investigated herbicides in sandy aquifers under natural flow conditions.

[Comparative studies of the filtration behavior of bacteria and organic particles in porous groundwater conductors. Fundamentals and methods]. / Vergleichende Untersuchung zum Filtrationsverhalten von Bakterien und organischen Partikeln in Porengrundwasserleitern. I. Grundlagen und Methoden.

In subsurface aquatic environments two groups of micro-organisms are observed: allochthonous bacteria and viruses, which as contaminants are eliminated from water after some time, and autochthonous groundwater micro-organisms, which belong to the natural subsurface environment and may reach very high abundances under favourable conditions--especially in the presence of a high nutrient supply. The migration of micro-organisms is controlled by flow length dependent transport processes (advection--dispersion, adsorption--desorption), and predominantly by filtration. This can be described on the basis of an expanded advection--dispersion concept. The filter effects in a certain porous aquifer can be quantified by the filter efficiency (filter factor) as a measure of the specific decrease of an initial concentration on a certain flow length. Recent laboratory experiments show that for sand the filter factor depends on the respective microbial species and is highly correlated to the effective grain diameter of the porous material, which is routinely determined in hydrogeology. Experiments with columns filled with quartz sand using the bacteria species Escherichia coli ATCC 11229, Pseudomonas cepacia DSM 50181, Streptococcus faecalis ATCC 6569, and polystyrene beads with similar density and diameters show that the filter factor is controlled by the grain size of filter material, the flow velocity, the diameter of the particle and the ionic strength of the water: The filter factor is specific for each microbial species for the same conditions of the aquatic environment. The filter factor decreases one order of magnitude if the flow velocity increases in the same order. A major control of the filter factor is the grain size. Conventionally the grain size is used as characteristic length instead of the pore size which, although it should be the real reference date, is relatively difficult to measure. For the assessment of the filter factor, the grain size d10, taken from the grain size distribution curve, can be used. The influence of the particle diameter on the filter factor, which was predicted by the filtration theory, was confirmed. The minimum values of the filter factor were encountered at particle diameters of about 1 micron, which is about the size of bacteria. The filter factor is influenced strongly by the ionic strength in water with low ionic strength, whereas in water of higher ionic strength its influence can be neglected. These relationships can be formulated into empirical equations, which allow prediction of the filter factor for given hydraulic conditions.