S U M M A R YA LArge Reservoir Simulator (LARS) was equipped with an electrical resistivity tomography (ERT) array to monitor hydrate formation and dissociation experiments. During two hydrate formation experiments reaching 90 per cent bulk hydrate saturation, frequent measurements of the electrical properties within the sediment sample were performed. Subsequently, several common mixing rules, including two different interpretations of Archie's law, were tested to convert the obtained distribution of the electrical resistivity into the spatial distribution of local hydrate saturation. It turned out that the best results estimating values of local hydrate saturation were obtained using the Archie var-phi approach where the increasing hydrate phase is interpreted as part of the sediment grain framework reducing the sample's porosity. These values of local hydrate saturation were used to determine local permeabilities by applying the Carman-Kozeny relation. The formed hydrates were dissociated via depressurization. The decomposition onset as well as areas featuring hydrates and free gas were inferred from the ERT results. Supplemental consideration of temperature and pressure data granted information on discrete areas of hydrate dissociation.

Gas hydrate production is still in the test phase. It is only now that numerical models are being developed to describe data and production scenarios. Laboratory experiments are carried out to test the rationale of the conceptual models and deliver input data. Major experimental challenges include (I) the simulation of a natural three-phase system of sand−hydrate− liquid with known and high hydrate saturations and (II) the simulation of transport behavior as deduced from field data. The large-scale reservoir simulator (LARS; 210 L sample) at the GFZ has met these challenges and allowed for the first simulation of the gas production test from permafrost hydrates at the Mallik drill site (Canada) via multistage depressurization. At the starting position, hydrate saturation was as high as 90%, formed from dissolved methane only. Whereas gas hydrate dissociation determined the flow patterns in the early pressure stages, the importance of different transport behaviors increased at lower pressure stages and increasing water content. Gas flow patterns as observed in Mallik were recorded. While the conceptual model for the experimental data does agree with the model proposed for Mallik at moderate and low gas production, it is different at high gas production rates.

Figure 3. (A) Progress of pressure (upper panel) and temperature (lower panel) during experiment A. For the location of the temperature sensors, see upper left corner of Figure 3B. (B) P−T paths as recorded during experiment A in comparison to calculated methane hydrate stability curves at 9.5 g/dm 3 NaCl (using CSMGem software). 3 The red arrows indicate the induced pressure steps, and the dashed blue lines indicate the accompanying average temperature changes. The processes are numbered consecutively. (C) Progress of pressure (upper panel) and temperature (lower panel) during experiment B with sensor distribution in Figure 3D. (D) P−T paths as recorded during experiment B. The measurement distance is 5 and 60 s in experiments A and B, respectively.

In the northern Sierra of Peru, water scarcity issues arise owing to the seasonal rainfall distribution and the lack of appropriate natural water storage capacity of river basins. The present study assesses the base flow and water storage volume of the Ronquillo watershed, an important rivulet for water abstraction for the city of Cajamarca. Mean base flow is 184 ls -1 , thus representing 44 percent of total stream flow. Flow recession curve analysis yields a mean catchment water storage volume of 3.57 × 10 6 m³, which corresponds to a runoff depth of 85 mm. The dischargeable water storage volume of Andosols, a soil type known to be a very important water reservoir in the Andes, corresponds to a runoff depth of 33 mm. Moreover, the study shows that the geological environment is of major relevance. Springs (18 mm) and an effluent flow regime (20 mm) contribute significantly to dry seasonal runoff. The findings imply that water conservation in the Ronquillo watershed should place emphasis not only on the preservation of soils, but also on subsurface water flow paths, as water availability is affected by processes operating beyond topographically derived catchment boundaries. En el norte de la Sierra del Perú, surgen problemas de escasez de agua debido a la distribución de la precipitación estacional y la falta de capacidad de almacenamiento de agua natural adecuado de las cuencas hidrográficas. El presente estudio evalúa el flujo base y el volumen de almacenamiento de agua de la cuenca del Ronquillo, un arroyo importante para la extracción de agua para la ciudad de Cajamarca. El flujo media de base es de 184 ls -1 , lo que representa 44 porciento del flujo total de corriente. Del análisis de la curva de flujo de recesión se obtiene un volumen de almacenamiento de agua de captación media de 3.57 × 10 6 m³, lo que corresponde a una profundidad de 85 mm de escurrimiento. El volumen de almacenamiento de agua descargable de Andosoles, un tipo de suelo conocido por ser un depósito de agua muy importante en los Andes, corresponde a una profundidad de 33 mm de escurrimiento. Por otra parte, el estudio muestra que el entorno geológico es de gran relevancia. Ojos de agua (18 mm) y un régimen de caudal de los efluentes (20 mm) contribuyen de manera significativa a secar la escorrentía estacional. Los resultados implican que la conservación del agua en la cuenca del Ronquillo debe poner énfasis no solo en la conservación de los suelos, sino también en las trayectorias de flujo de agua del subsuelo, ya que la disponibilidad de agua se ve afectada por los procesos que operan más allá de las fronteras de captación derivadas topográficamente.

At the GFZ German Research Centre for Geosciences we have developed a safe and efficient method which allows for the decomposition of gas hydrates by the supply of heat inside the reservoir. The heat is generated in situ by a catalytic combustion of methane in a counter-current heat-exchange reactor. The reactor that Rudy Rogers, Professor Emeritus in Chemical Engineering at Mississippi State University, referred to as the "Schicks Combustor" is placed in a borehole in such way that the hot reaction zone is situated in the area of the hydrate layer. The counter-current heat-exchange reactor developed at GFZ generates heat via a flameless catalytic oxidation of methane at a noble metal catalyst. The system is closed i.e. there is no contact of the reactants, catalyst and environment. For safety reasons, methane and air are fed separately through a tube-in-tube arrangement into the mixing chamber. Due to its cooling effect and for safety reasons air instead of pure oxygen is used. From the mixing chamber the gas mixture arrives in defined quantities on the catalyst bed, where methane and oxygen are converted into carbon dioxide and water. The hot product gases release their heat via an aluminum foam to the outer wall of the reactor and then to the environment. Simultaneously, the incoming gases are preheated. The reaction runs stable and autonomous between 673 and 823 K. The counter-current heat-exchange reactor was designed as a lab reactor and a borehole tool. The lab reactor was tested in a reservoir simulator to investigate the heat transfer into gas hydrate bearing sediments. Therefore methane hydrate was generated in the LArge Reservoir Simulator (LARS), an autoclave with a volume of 425 L. In a test with 80% hydrate saturation, the reservoir simulator warmed up within 12 hours after the ignition of the catalyst to such an extent that the temperature of the complete sample was above the dissociation temperature of the previously formed methane hydrate which dissociated completely and methane could therefore be produced. During this test, only 15% of the produced CH4 was consumed to generate the energy needed for the thermal dissociation of the hydrates. The experience with the laboratory reactor served as basis for the design of a borehole tool which is suitable for the application in natural gas hydrate reservoirs. The borehole tool has a total length of 5120 mm, an outer diameter of 90 mm and weighs ca. 100 kg. First results from field tests at the continental deep drilling site KTB in Windischeschenbach, Germany, confirm that the borehole tool reliably produces heat at depth.