Tracer test is an old technique originally developed to determine fluid flow connectivity in surface water paths and aquifers. In the last decades it has been greatly enhanced and is used in many earth science applications in hydrology, aquifers, mining, oil reservoirs, geothermal fields, pollutants dispersion, atmospheric circulation, oceanology, geology, etc. (Käss 1998). Presently, tracer tests constitute an important and decisive technique in determining communication channels in underground porous formations. It can provide fundamental information on the fluid motion, can determine formation properties and can offer solid elements when validating geological models. A traditional inter-well tracer test in underground formations consists in the release at a certain place (injection well) of a pulse of a chemical, biological or radioactive element, which is carried by the fluid following the flow trajectories and is later observed at other downstream sites (production or observation wells), normally at very low concentrations. The amount of tracer observed along the time (tracer breakthrough curve) provides therefore information on the flow pattern, communication channel characteristics and the porous media properties. In order to avoid the employment of huge amounts of a tracing product in tracer tests, those tracing elements that can be perfectly integrated in the fluid and can be detected at very low concentrations are preferred. Large amounts of a tracer are hard to manage and can modify the original system conditions. Therefore radioactive elements such asTritium in tritiated water or in tritiated methane are frequently used. Many analytical and numerical models to describe tracer test in Geohydrology and oil reservoir research have been developed in the last decades (see for exampleVan Genuchten 1981, Maloszewski & Zuber 1990, Ramírez et al. 1993, Zemel 1995, Charbeneau 2000), which describe diverse circumstances and porous media types. A very important situtation is when geological faults are present, which are structural features frequently appearing in aquifers, oil reservoirs, and geothermal fields. Faults can work as a transversal fluid flow barrier or oppositely, as high permeability longitudinal conduit; therefore, their presence can seriously impact fluid flow dynamics. In oil reservoirs faults can dramatically reduce the efficiency of secondary and improved recovery procedures, and can importantly increase the undesired transport of pollutants in aquifers. The fault structure regularly consists of a central low permeability slab-like core, flanked by a high-permeability damaged zone at its both sides, which contains an intricate network of fractures with different sizes and characteristics. The fluid conduction properties of a fault depend on the composition and thickness of these damage zones (Caine et al. 1996, Fairley & Hinds 2004). For example, when conductive faults are present in a reservoir under secondary or improved oil recovery processes, conductive faults can seriously reduce the swept efficiency as a consequence of fluid oil-by-passing situations. The disposal of high radioactive waste in deep repositories located in fractured formations can represent a safety hazard. Also, contaminant spills on fields can rapidly pollute aquifers if underground conductive faults are present. Thus, knowing the presence of faults and their fluid conduction characteristics is of great relevance. Tracer tests can play a decisive role in determining those characteristics. To this purpose mathematical models

98 Mathematical and numerical modeling in porous media

to interpret field tracer data are necessary. Diverse models to describe tracer transport in a fracture network have been developed in order to treat tracer flow in faults (see for example Sudicky & Frind 1982 and Maloszewski & Zuber 1985) but they did not consider an important issue in oil reservoirs and geothermal fields, which is that injection and production wells can be located outside the fault. Recently some analytical tracer test models have been developed to treat this situation (Coronado & Ramírez-Sabag 2008, Coronado et al. 2010). An important point in these models is the introduction of three coupled tracer path regions, which represent the zone from the injection well to the fault, the path inside the fault self, and the region from the fault to the production well. In one of these works (Coronado et al. 2010) a successful application of the models to field data is made, through them the orientation of the fault could be calculated. Specifically, this is a very valuable information for geologists and reservoir engineers. In this work we present a complementary numerical model developed to describe a tracer test in geological formations with a conductive fault. The results from the analytical models are compared against the numerical model results, as a way to validate the structural model consistency. The numerical model has been set up by using the worldwide known simulator UTCHEM (Delshad et al. 1996).