Simulating the in situ physical properties of the upper Muschelkalk aquifer

The Swiss Molasse Basin (SMB) and the adjacent Folded Jura contain the most suitable deep aquifers for storage in Switzerland (Chevalier et al., 2010). Due to the unsuccessful petroleum exploration in the SMB and Jura, the basin has not been industrially developed and knowledge of the sub-surface is limited.

The upper part of the Muschelkalk has the strongest potential for CO2 storage, due to its favorable properties. This unit consists of porous dolomites in upper half (Trigodonusdolomit), and limestones in the lower half of the unit (Hauptmuschelkalk). In addition to this, Chevalier et al. (2010) discussed that the salinity of the Muschelkalk aquifer is highly variable, which could affect the CO2 storage properties. Since the Gipskeuper, or the seal, is composed claystone and anhydrite rock sequence, with low permeability, they represent compositions of a good and extensive seal.

Anthropogenic emission of carbon dioxide (CO2) is widely believed to contribute to climate change. Hence, it has been recognized that there is a need for the reduction of carbon dioxide emissions. Recent technologies for carbon capture and storage, also known as CCS, have therefore been developed. CCS has been investigated in different parts of the world, and is currently under investigation in Switzerland (Chevallier et al., 2010). In this study, the aim is the characterization of one potential storage aquifer and seal pair, in order to gain a better idea of where and how deep to store CO2 in Switzerland. The potential units comprise the Gipskeuper (seal), and the underlying upper Muschelkalk (aquifer). The aquifer is divided into the Trigodonusdolomit and the Hauptmuschelkalk units. A series of laboratory measurement techniques were employed to characterize the physical properties of core samples of dolomite, recovered from drill core. These measurements consist of density, porosity, permeability and ultrasonic wave velocities. Specifically, the permeability and ultrasonic velocity measurements were performed at conditions similar to reservoir conditions, with elevated pressure, spanning near surface conditions to >3 km depth. The permeability is one of the key parameters in this project since the well logging data can only extract a very rough permeability and hydraulic conductivity data by injection pump tests. The laboratory measurements were compared with well logging data (courtesy of the National Cooperative for the Disposal of Radioactive Waste, Nagra), from the Benken and Weiach drill sites (northern Switzerland) where the core was recovered. Of particular interest is to understand and quantify the distribution of the physical properties, with the aim to better understand the geophysical signature of the Muschelkalk aquifer (in terms of borehole properties, and seismic surveys).

(The study found that, in general, the measured velocities show a positive correlation with density and an inverse correlation with porosity. It was also found that higher porosity tends to reduce the ultrasonic velocity. A simple but effective model is created to predict the seismic reflection coefficient (R) in the Trigodonusdolomit, with porosity varying from 0 to 50 %. At low porosity, the reflection expected at lithological boundaries is weak, but with increasing porosity in Trigodonusdolomit, the amount of reflected energy increases. At the Gipskeuper to Trigodonusdolomit horizon the reflection coefficient is negative. In contrast R is positive at the Trigodonusdolomit to Hauptmuschelkalk horizon. The empirical model agree qualitatively well with the reflection coefficient observed in seismic surveys at the Benken and Weiach drill sites (Nagra, 1989; 2001). Differences in the absolute values of the reflection coefficient are expected, likely related to the scale differences in laboratory and borehole logging results, compared to seismic surveys. In addition, although the volume of pores is the main contributing factor to reduction in seismic velocity, other factors need to be taken into consideration, especially the pore structure resulting from diagenetic alteration. Nevertheless this model provides a preliminary tool to understand the seismic reflection due to the porous Trigodonusdolomit.)