The study of air-core dynamics in hydrocyclone operation is essential for predicting separation efficiency in mineral processing. Electrical resistance tomography (ERT) allows non-invasive measurements of hydrocyclone internal flow dynamics. Existing regularisation methods such as Gauss-Newton and TV norms rely on optimisation techniques which have very high computational cost for very fine meshes. In contrast, the monotonicity method has much lower computational cost as it does not solve such minimisation problems. In this paper, results from the monotonicity method, Gauss-Newton (GN) (Tikhonov) and Total-Variation (TV) norm-based minimisation methods for estimating the air-core diameter in hydrocyclones are compared. We study these methods on the laboratory hydrocyclone experimental two-phase flow data. The results obtained are validated against High-speed video (HSV) imaging and Computational Fluid Dynamics (CFD) results. The results on different mesh sizes, in particular for very fine meshes with the number of cells ranging from 30 000 to above 100 000 are also investigated. The data used in this study is obtained from ITS Z8000 ERT system. This system does not measure the voltages on the current carrying electrodes. Since these voltage values are required for the monotonicity method, we implement an interpolation method to compute them. The results show that the monotonicity method provides a closer approximation of the interface between air and the water phases.