Smart cement modified with iron oxide nanoparticles to enhance the piezoresistive behavior and compressive strength for oil well applications

摘要

In this study, smart cement with a 0.38 water-to-cement ratio was modified with iron oxide nanoparticles (NanoFe(2)O(3)) to have better sensing properties, so that the behavior can be monitored at various stages of construction and during the service life of wells. A series of experiments evaluated the piezoresistive smart cement behavior with and without NanoFe(2)O(3) in order to identify the most reliable sensing properties that can also be relatively easily monitored. Tests were performed on the smart cement from the time of mixing to a hardened state behavior. When oil well cement (Class H) was modified with 0.1% of conductive filler, the piezoresistive behavior of the hardened smart cement was substantially improved without affecting the setting properties of the cement. During the initial setting the electrical resistivity changed with time based on the amount of NanoFe(2)O(3) used to modify the smart oil well cement. A new quantification concept has been developed to characterize the smart cement curing based on electrical resistivity changes in the first 24 h of curing. Addition of 1% NanoFe(2)O(3) increased the compressive strength of the smart cement by 26% and 40% after 1 day and 28 days of curing respectively. The modulus of elasticity of the smart cement increased with the addition of 1% NanoFe(2)O(3) by 29% and 28% after 1 day and 28 days of curing respectively. A nonlinear curing model was used to predict the changes in electrical resistivity with curing time. The piezoresistivity of smart cement with NanoFe(2)O(3) was over 750 times higher than the unmodified cement depending on the curing time and nanoparticle content. Also the nonlinear stress-strain and stress-change in resistivity relationships predicated the experimental results very well. Effects of curing time and NanoFe(2)O(3) content on the model parameters have been quantified using a nonlinear model.