A novel predictive model for the compressive strength development in class G cement slurries at different temperatures, retarder, and salt concentrations

Using oil well cement with high early-age compressive strength plays a crucial role in securing the long-term integrity of the wellbore. Hence, an accurate evaluation and estimation of delay in cement compressive strength becomes considerably important to reduce cost expenditure in drilling operations. In this research, several experiments were run utilizing a non-destructive approach by using the Ultrasonic Cement Analyzer (UCA) device to evaluate the influence of temperature, retarder, and salt concentrations on the delay of 24-hour cement compressive strength. Moreover, a computer-based Radial Basis Function optimized by Particle Swarm Optimization (PSO-RBF) model with three layers was developed for the estimation of experimental compressive strength data. In addition, a rate-constant model was also developed for the prediction of the measured compressive strength profile of cement slurries. The results showed that increasing the temperature decreases the delay in compressive strength. Also, increasing the retarder and salt concentration in cement slurry can increase the delay in compressive strength. The modeling results revealed that the overall correlation coefficient (R 2 ) and Average Absolute Relative Deviation (AARD%) of the PSO-RBF model were 0.9999 and 3.36 %, respectively. The developed models are useful for rapid prediction of the compressive profile of cement slurries when there is no or a limited number of experimental data.