Nanofluids are colloidal suspensions constituted of nanoparticles and typical heat transfer fluids which have shown potential in yielding enhanced heat transport for many applications. Significant attention has been paid to their thermal conductivity enhancement which has been alleged, in some cases, to exceed theoretical limits classifying the enhancement as “anomalous”. The present study aims to quantitatively investigate the nature of the enhancements reported in the literature and classify their alignment with theoretical predictions. To do so, a rigorous and objective mathematical analysis method has been employed. The novelty and value of the present work lies in the deeper characterisation and understanding of the anomalous observations reported. The present analytical study focuses on (spherical) Al2O3–water nanofluids. It was discovered that studies involving low nanoparticle concentrations ($$\phi $$ ϕ ≤ 0.2 vol%) and the use of electrostatic stabilisation (through pH control) as opposed to steric stabilisation (using surfactants) as suspension stability control methods are likely to report anomalous effects. An exceptional case was observed for d < 15 nm, where to achieve anomalous enhancement, surfactants and pH controllers should not be used to prevent significant interfacial resistance. The shared characteristics of these anomalous observations indicate that nanofluid preparation effects are linked to the underlying physical mechanisms of heat transfer involved and those should be further investigated. The failure of studies attempting to replicate anomalous thermal conductivity enhancement in the literature could hence be understood, as these did not satisfy the conditions required to lead to an anomalous enhancement. The role of measurement errors was also considered.

Stability is key to sustaining the colloidal properties of nanofluids and by extension the beneficial thermophysical properties they exhibit for practical applications. Nanofluid suspensions are typically prepared through ultrasonic dispersion of nanoparticles, either using low-power ultrasonic baths or high-power ultrasonic probes. It has been observed that high-power probes, although achieving nanoparticle size reduction in a short time, can also cause considerable aggregation of particles and hence reduction in colloidal stability with excessive application. This effect is not observed in low-power ultrasonic dispersion applications. This discrepancy and its sources are explored and explained in the current chapter, through consideration of particle breakup mechanisms (fragmentation versus erosion) and the fusion of particles due to high-velocity interparticle collisions. Stability is known to be linked to solution pH; for example a pH value far from the isoelectric point yields a surface charge in the dispersed phase, which enhances stability through coulombic repulsion. Ultrasonication has been observed to affect the pH of nanofluid solutions. High-power devices are unable to affect pH change in dilute alumina–water nanofluids (ϕ &lt; 0.01 vol%), whereas low-power devices can. This is hypothesised to be due to the dominant breakup mechanism, i.e., erosion in low-power baths versus fragmentation in high-power probes. Hence, to improve nanofluid stability, it is recommended to use low-power sonication where possible, and source nanoparticles in aqueous form. If a high-power ultrasonic probe must be used, the duration and amplitude should be reduced to avoid the induction of significant stability reduction.