Robust predictive models of dynamic bubbly systems of nanoparticle-liquid mixtures are vital to the design and assessment of relevant industrial systems. Previous attempts to model bubbly flows of dilute nanofluids using the classic two-phase flow models were unsuccessful although the apparent hydrodynamic properties of the dilute nanoparticle-liquid mixtures were only negligibly different to those of their pure base liquids. Emerging studies demonstrated that when bubbles exist in the mixture, nanoparticles tend to spontaneously aggregate at the bubble interface, forming a layer of "colloidal armour" and making the bubble interface partially rigid and less mobile. The colloidal armour also significantly modifies the characteristics of the bubble-liquid and bubble-bubble interactions. Therefore, it was proposed that the key job when developing a predictive model based on classic two-phase flow models is to reformulate the bubble-liquid and bubble-bubble interactions. However, the adsorption of nanoparticles in dynamic bubbly systems has rarely been studied. The lack of mechanistic understanding has severely hindered the model development. Therefore, this study reviews the common findings yielded from experimental and numerical investigations reported in literature, with the aim to clarify the critical points to address when modelling bubbly flows containing nanoparticles using the classic two-fluid and MUSIG models. (C) 2017 Elsevier Ltd. All rights reserved.