Summary A cement crack is a typical cause of oil and gas well failure. Cracks weaken cement, reducing zonal isolation and fluid leakage. Nanoparticle (NP) gels are being tested for fracture treatment. When crushed into cracks, the flow behavior of NP problem solutions should be predicted. The potential efficacy of utilizing NP gels as a remedial measure for fractures is currently under investigation. It would be advantageous to determine if the flow behavior of solutions for NP problems can be anticipated when they are compressed into crevices. This study aimed to analyze the behavior of nano-silica solutions as they flow through ducts with rectangular cross-sections and varying crack dimensions. The introduction of NP solutions into the core leads to a decrease in pressure, which suggests that the nano-silica has been effectively transported through the crack. As the size of the fracture decreases, there is a corresponding increase in pressure drops, while the flow rate experiences a concurrent increase. This study presents responses of a pressure gradient to fluid concentration for a range of fracture widths, heights, and flow rates. The prediction of laminar flow in ducts is based on the linear correlation between the flow rate and the pressure gradient. Furthermore, the reduced pressure gradient indicates enhanced fluid flow within the fracture because of the amplified slot width. The fluid flow model proposed by Guo et al. (2022) was utilized to conduct a comparative analysis with the experimental data. Compared with test data, the model differs by roughly 90%. The technical cause of the flow model-observed data discrepancies is unknown. The flow model did not account for friction between NPs-NPs and NPs-walls in rough ducts. An empirical correlation has been found that quantifies the ratio as a function of nonsilica solution flow rate, cross-sectional geometry parameters, and nano-silica concentration. The correlation was calculated using nonlinear regression. The empirical relation and actual ratio have a significant correlation, as shown by R2 = 0.8965. In practice, Guo et al.’s (2022) hydraulic model’s pressure drops should be multiplied by the empirical correlation’s ratio to reduce errors.