Heat loss from vertical solar hot water storage cylinders is detrimental to their thermal performance. As such, many researchers devoted their attention to investigating changes in the rate of heat loss in response to different cylinder designs and initial temperature profiles. Despite the extensive literature relating to heat loss from vertical solar hot water storage cylinders, few studies have examined passive means of controlling the natural convection flow behavior inside the tank as it loses heat to the surroundings.

Based on this gap in our understanding, a preliminary study was performed to examine transient natural convection inside vertical hot water cylinders experiencinga standing heat loss using Computational Fluid Dynamics (CFD). The model was validated with experimental temperature measurements and boundary layer velocity measurements from Particle Image Velocimetry (PIV). From this, it was found that the natural convection heat transfer coefficient changed with time, partly due to the development of a thermocline in the tank.

Based on this observation, it was decided to decouple the natural convective heat transfer coefficient from time and the temperature profile in the cylinder by modellingit under a quasi-steady state condition. Using this approach, it was shown that natural convection heat transfer decreases with increases to the cylinder’s aspect ratio. Furthermore, it allowed a generalized heat transfer correlation to be developed to describe the natural convection heat transfer inside the cylinder regardless of time or the temperature profile.

Over the course of these preliminary studies, it was apparent that the side wall boundary layer flow had a major impact on the natural convection heat transfer. As such, research was focused on attempting to control this flow. In the first instance, a series of cylindrical jacket-type baffles were placed inside the cylinder. It was shown how the baffle jacket was able to deliver a 40% reduction in the Nusselt number and led to an understanding of the baffle geometry best suited to this purpose. This led to the development of a generalized heat transfer correlation to determine Nusselt number inside tanks with baffle jackets as a function of the baffle’s geometry and proportions.

Given the potential for simple baffle structures to reduce heat loss, it was decided to examine if another modification could also deliver an improvement: replacing the usual insulation material with a layer of trapped air (much like an inverted ThermosTM flask). Again, CFD was used to examine the flow in ‘trapped’ air gaps around the cylinder of between 10 and 100 mm. It was shown that increasing the aspect ratio of the inner cylinder (i.e. the storage tank) alters the flow behavior in the region intersecting the side and top air gap cavities. Because of this, convective heat transfer decreases as the aspect ratio increases. Based on the analysis, a heat transfer correlation that describes the change of the convective heat transfer coefficient with geometrical changes to the ‘trapped’ air insulation layer was developed.

Finally, using the correlations developed in this work, the long-term performance of a solar water heating system incorporating the studied storage cylinders was modelled analytically. The results indicate that a baffle jacket could improve the thermal performance by up to 3%, while the level of insulation provided by an air cavity could enhance the thermal efficiency by up to 8%. This indicates that the amount of traditional insulation required could potentially be reduced by using an air insulation layer.

In summary, this work has delivered an improved relationship for predicting the rate of natural convection heat transfer from vertical solar hot water storage cylinders, by developing a generalized correlation that can predict it irrespective of time or the initial temperature profile. Furthermore, the work has shown that the use of a simple passive baffle that alters the natural convection flow inside a cylinder can reduce the heat loss in experiences. Finally, it was also found that, with good design, an air gap that serves as an insulation layer could also markedly reduce heat loss. It suggests that despite the mature nature of the technology, there may be opportunities to include simple design modifications that could lower capital costs without compromising the system’s performance.