Geochemical modeling of CO2 injection and gypsum precipitation at the Ketzin CO2 storage site

Abstract

Gypsum crystals are found at the well perforation of observation well Ktzi 202 of the test site for CO2 storage at Ketzin, Germany. XRD analysis confirms pure gypsum. Fluid samples before and after CO2 injection are analyzed. Geochemical modeling is conducted to identify the mechanisms that lead to gypsum formation. The modeling is carried out with PHREEQC and Pitzer database due to the high salinity of up to 5 mol per kg water. Due to their significantly higher reactivity compared to other minerals like silicates, calcite, dolomite, magnesite, gypsum, anhydrite, and halite are considered as primary mineral phases for matching the observed brine compositions in our simulations. Calcite, dolomite, and gypsum are close to saturation before and after CO2 injection. Dolomite shows the highest reactivity and mainly contributes to buffering the brine pH that initially decreased due to CO2 injection. The contribution of calcite to the pH-buffering is only minor. Gypsum and anhydrite are no geochemically active minerals before injection. After CO2 injection, gypsum precipitation may occur by two mechanisms: (i) dissociation of CO2 decreases activity of water and, therefore, increases the saturation of all minerals and (ii) dolomite dissolution due to pH-buffering releases Ca2+ ions into solution and shifts the mass action to gypsum. Gypsum precipitation decreases with increasing temperature but increases with increasing partial CO2 pressure. Our calculations show that calcium sulfate precipitation increases by a factor of 5 to a depth of 2000 m when Ketzin pressure and temperature are extrapolated. In general, gypsum precipitation constitutes a potential clogging hazard during CO2 storage and could negatively impact safe site operation. In the presented Ketzin example, this threat is only minor since the total amount of gypsum precipitation is relatively small.