Prediction of fluid-flow processes involving heat transfer can be accessed via two differing paths; experimental investigation and theoretical calculation. The engineering problem presented in this work is governed not only by a desire to optimise the supply of solar heat to a Thermal Energy Store (TES) by means of Immersed coil Heat Exchangers (IHX); the performance of an isolated heat exchanger, perhaps reducing its cost of manufacture, or matters of the system performance as a whole (effects of stratification) - but also to understand intimately the overarching physics at play. As clever as the tools of our investigation seem, what is certainly of more importance is the way we interpret results gained from them.

The most reliable information we can obtain is from direct experimentation, however full scale testing at the desired resolution of system parameters, geometrical variations and real conditions is prohibitively expensive and time consuming. The scaling, reduction and separation of tests into isolated operations allow a significant saving on this but our ability to extrapolate from them is limited. Neither is the equipment necessary for experimentation free from error. A look at a classical heat transfer or fluid mechanics textbook reveals that only a small set of practical problems can be solved in a closed mathematical form, and that even these solutions can present a formidable task. Thus far development in numerical methods has shown great promise in that the implications of a mathematical model can be worked out for almost any application under ideal or real conditions, at a low cost and high resolution. These implications, however, do not necessarily imply reality itself.

A review of literature reveals a plethora of dissertations, articles and proceedings significant to the common understanding, modelling and conception of our problem at hand. The findings contained in these works have had a large influence on the direction this work has taken; in particular, observations made, whether in theory, numerical modelling or experiments, served as a basis to be built upon.

As early as 1989 [15] it was shown that an optimal exergetic utilisation of incoming energy requires as high an overall heat transfer coefficient (U-value) as possible and a generous charging time period (two hours) for a given mass-flow. The simple reason behind this being that the exergetic efficiency is inverse proportional to the mass-flow; the lower the flow, the higher the efficiency and longer the filling time. The dissertation from Messerschmid [27] took up this criteria and, seeking to optimise the geometric parameters of an IHX with mass-flows of 1000– 2000 l.h⁻¹, found that the most important parameter is that of the relationship from coil-helixpitch to tube diameter, and that the distance between tube helices should be about the same as the diameter of the tube. Other works exhibit similar findings [1, 41, 33] without specifying what an optimum might be.

Along a path of investigation where the operating conditions reflect the charging stores with solar thermal heat, the phenomena of stratification (first published 1979 [40]) has come to be known as a criteria of at least, when not more importance [36]. Significant efforts have been made in defining criteria that capture differing degrees of stratification efficiency [16] and evidence for optimising immersed coil heat exchangers under stratified and low collector flow conditions can also be found [39].

However, experiments in the past and also in the investigations presented here have shown that stratification is hard to achieve when charging with an IHX. Thus the questions connected in the topic of optimising the charge of a TES by means of an IHX are manifold:

- can the heat exchanger efficiency be improved significantly by changing its geometrical parameters (while keeping the same costs),

- can the storage stratification be improved significantly by changing the geometrical parameters of the IHX,

- can the system efficiency be improved significantly by changing the geometrical parameters of the IHX,

- can the layout of an IHX be improved for solar thermal systems in order to optimise the combination of costs and system efficiency?

Paradoxically these questions do not necessarily overlap, nor even guide the optimisation process in the same direction. The physical understanding of each component described independently through measurement and the effect of their combination in a system with ancillary components is complex and hard to generalise.

From a simple heat transfer formulation in Section 2 to detailed experimentation in Section 3.7 and numerical studies utilising established Computational Fluid Dynamics (CFD) software in Section 3.9, this report seeks to commiserate the knowledge from both paths of inquisition in a systematic and holistic approach.