Pore scale assessment of disrupted hydrochemical equilibria in geothermal systems

Abstract

Many geothermal plants in the north alpine foreland basin (NAFB) are affected by the precipitation of calcium carbonate and struggle with efficiency losses and sometimes safety problems. The adaption and implementation of predictive maintenance strategies relies on the accuracy of the prediction and make sense if the costs of early maintenance are significantly less compared to the costs of a replacement of failed parts. Therefore, the accurate description of the kinetics of inorganic precipitations have to be extended to include the fluid dynamics and the interaction of the precipitates with different materials used in the geothermal cycle. The experimental concept also applies to fluid-rock interactions which can alter the properties of the reservoir. The parametrization of transferable hydrogeochemical models performs best with data on a single interface, single crystal level. This data allows to elucidate the underlying process kinetics and improve existing strategies. The combination of Raman spectroscopy and Quartz crystal microbalance (QCM) opened a way to quantify the formation of carbonate precipitates. Here, the QCM can measure the total mass of attached particles while Raman microscopy identifies the crystal polymorph. Running these experiments in microfluidic channels allows to assess the effects of physical stress on the formation and to test inhibition and removal stratgies. The QCM sensor was placed in a microfluidic channel and tap water (carbonate rich, Ca2+ concentration approx. 2.25 mM, pH approx. 7.50) and sodium hydroxide (0.10 M) were injected through the two inlet channels. As the lime carbonic acid equilibrium shifts due to the pH increase, precipitations are formed. The adhesive forces on different materials (SiO2, aluminium, steel) were studied by changing the flow velocities and chemical cleaning processes were mimicked by injecting an acid (HCl). The precipitation of CaCO3 on the QCM sensor (sensitivity in the low mg-range) was less than 20% of the theoretical amount. This underlines the importance to include the fluid dynamics into the assessment models. The experimental data under dynamic conditions was modelled with a combination of CFD simulations with PhreeqC: The particle flow velocity and the precipitates formed depend on the depth as well as local equilibrium changes. The preferred location of scaling could be adequately simulated. This quantification of the effects of the shear stress and the material properties on the scaling efficiency is a further step towards predictive maintenance strategies and a solid comprehensive site assessment during the planning stage, which should improve the sustainability and the much-needed attractiveness of this energy sector.