Zusammenfassung
The storage of carbon dioxide (CO2) in geological formations is a promising option to reduce emissions of greenhouse gases to the atmosphere. During the proposed CO2 injection process, application of suitable techniques for monitoring of the induced changes in the subsurface is required. Existing models for the spreading of the CO2, as well as displacement processes or resulting issues from mutual solubility between brine and CO2 associated with saturation changes, need to be checked.
For well logging in cased boreholes only a limited number of techniques such as radiometric pulsed neutron-gamma (PNG) logging are applicable. The main PNG derived parameters are the macroscopic capture cross section (Σ) and the neutron porosity (TPHI) of the formation. The high measurement contrast between brine and CO2 results in a high sensitivity to evaluate saturation changes.
Previously, the conventional PNG saturation model based on a displacement process has been used for PNG interpretation in different CO2 storage projects in saline aquifers (e.g. Murray et al., 2010; Müller et al., 2007; Sakurai et al., 2005; Xue et al. 2006). The displacement saturation model accounts for mixing of the fluids in the pore space but ignores any mutual physico-chemical interaction. In addition to the displacement process, the mutual solubility between brine and CO2 adds further complex processes such as evaporation and salt precipitation.
Evaporation and precipitation processes are relevant in the vicinity of an injection well, where dry CO2 enters the reservoir. Moreover, modeling results show that gravity-forced upflow of CO2 and capillary-forced backflow of brine toward the injection point can also affect saturation changes and salt precipitation (e.g. Pruess und Müller 2009). The Σ brine value depends strongly on the brine salinity e.g. its chlorine content which makes PNG measurements suitable for evaporation and salt precipitation monitoring. Until now evaporation and precipitation processes are not considered in PNG saturation models.
For this purpose an extended PNG saturation model for NaCl-brines is developed. The extended PNG saturation model includes both the displacement and evaporation/precipitation processes weighted by their affected porosity proportions. Two scenarios are distinguished. If the repeat Σ value is smaller than the baseline Σ value, first the displacement saturation model and subsequently the evaporation/precipitation saturation model are applied to the affected porosity proportions. If the repeat Σ value is greater than the baseline Σ value, the pore fluid Σ value is increased corresponding to an increased salt load due to capillary effects. For this purpose, TPHI can be used to derive CO2 saturation and Σ is used to derive brine and/or halite saturation.
Ketzin, the only German pilot test site for CO2 storage in saline aquifers, has one injection and two observation wells, where an extensive PNG monitoring program is performed. For the observation well, which is further away from the injection well, the conventional displacement saturation model is valid. In contrast, the observation well closest to the injection well shows that, in addition to displacement, the evaporation/precipitation process may have occurred in some depth intervals. For the injection well, both scenarios occur. The halite saturation distribution in the injection well seems to be controlled by the combined effects of changing injection rates associated with changing brine levels, lithology heterogeneities, and different saturation changing processes such as evaporation/precipitation or capillary forces.
Such salt precipitations were previously not detected in situ in connection with CO2 injection in saline aquifers. In general, it can be said that PNG monitoring in combination with the extended saturation model is suited to determine displacement and evaporation/precipitation processes for CO2 storage operations. The extended saturation model is needed, especially for injection wells. Otherwise the CO2 saturation will be erroneously underestimated.
The depleted gas field at the Altmark site, Germany, has been proposed as a suitable geological formation for CO2 storage operations in combination with enhanced gas recovery. The PNG monitoring potential, with considering of displacement, evaporation/precipitation processes and changing gas-water contacts, is evaluated. As a result, PNG logs can be applied for determination of changes in brine saturation and water content associated with salt precipitation, whereas changes in the gas composition are below the detection limit.
