The interest in modeling geochemical reactions has increased significantly for different improved oil recovery processes such as Alkali-Surfactant-Polymer (ASP) flood, low salinity water flood, and EDTA injection as a sacrificial agent in hard brine. Numerical simulation of multiphase flow coupled with geochemical reactions is challenging due to complex and coupled aqueous, aqueous-solid, and aqueousoleic reactions. These reactions have significant impact upon oil recovery, and hence a robust geochemical simulator is important.UTCHEM is a chemical flooding reservoir simulator with geochemical modeling capability. Nevertheless, one major limitation in the geochemical reactive engine of UTCHEM is assuming the activities of reactive species are equal to unity. In fact, the activity coefficients are strongly non-linear functions of the ionic strength of solution. One approach to tackle this deficiency was to couple UTCHEM (flow and transport) with IPhreeqc (a geochemical reactive engine). However, the simulator proved to be computationally expensive. Therefore, it is desirable to improve the geochemical reactive engine within UTCHEM.This paper presents the improvement of the geochemical reactive engine in UTCHEM including implementing different activity coefficient models for different reactive species, cation exchange reactions, and numerical convergence. Certain unknown concentrations are eliminated from the elemental mass balance equations and the reaction equations to reduce the computational burden. The Jacobian matrix and RHS are updated accordingly in the Newton-Raphson method for performing the batch reaction calculation.A low salinity water flood case is presented to validate the updated UTCHEM against PHREEQC and UTCHEM-IPhreeqc. The simulation studies indicated that the updated geochemical simulator succeeds in tackling the inaccuracy concerned in the original UTCHEM. Also, the updated version is more efficient compared to PHREEQC and UTCHEM-IPhreeqc with the same degree of accuracy. The updated geochemical simulator is then applied to model an ASP coreflood, in which EDTA is used as a scarifying agent to chelate calcium and magnesium ions. The experimental data of pH, oil recovery, and pressure drop were successfully history matched with predictions of the effluent concentrations of calcium and magnesium ions. A synthetic 3D ASP pilot case is successfully simulated considering effects of acid equilibrium constant on oil recovery.