Chemical Composition-Driven Modeling of Rheological Properties in Sustainable Oil Well Cement Using Individual Oxides and Indices with Micro- and Nano-Silica Additives

Yield stress and plastic viscosity (PV) are key rheological parameters controlling the flow behavior, placement efficiency, and zonal isolation of oil well cement slurries. Maintaining these properties within API standards is essential for well integrity. This study investigates the effect of the chemical composition of cement, particularly silicon dioxide (SiO2), on yield stress across three systems: base cement, nano-silica (NS)-modified cement, and silica fume (SF)-modified cement. In modified systems, total SiO2 includes contributions from both cement and additives. A dataset of 224 entries was analyzed alongside 358 entries from the literature to examine the relationship between yield stress and PV. Four quartic regression models were developed using different chemical descriptors: individual oxides (IOs), alumina-ferric ratio (AFR), silicate-metallic ratio (SMR), and lime modulus (LM). Results show that in base cement, the IO model achieved the highest accuracy (R 2 = 0.98, RMSE = 2.49), whereas the LM model performed worst (R 2 = 0.80, RMSE = 8.88). For NS cement, IO, and AFR models showed strong performance (R 2 = 0.96), whereas all models performed similarly well for SF systems (R 2 approximate to 0.97). Overall, models based on IOs outperformed combined indices. These findings highlight the critical role of oxide composition, particularly SiO2, in controlling cement rheology and support incorporating detailed chemical parameters into predictive models of oil well cement performance.