Abstract
Phase separation in a direct contact thermal energy storage (DC-TES) system
is experimentally investigated in this thesis. The phases in question are a heat
transfer fluid (HTF) and a phase change material (PCM) which are brought in
direct contact as part of an efficient energy storage technology for thermal energy.
Being immiscible fluids, they create an emulsion consisting of a dispersed and a
continuous phase as a result of their mixing. Heat is exchanged efficiently between
the fluids due to the large interfacial area between them, which represents
the core benefit of a DC-TES. However, it is absolutely crucial for the system that
the phases are rapidly and completely separated once heat has been exchanged.
Incomplete phase separation and resulting transportation (carry-over) of the PCM
is currently a critical bottleneck for DC-TESs.
The DC-TES concept was first proposed in 1957 and has since been reported in
several investigations, mostly focused on its high thermal performance compared
to traditional thermal energy storage technologies. To date, however, no robust
solution has been demonstrated to address the phase separation issue. Solving
this problem requires in-depth knowledge of the multiphase flow dynamics within
a DC-TES, specifically the behavior and stability of the HTF/PCM emulsion.
In this work, an experimental DC-TES was developed at SINTEF’s Multiphase
Flow laboratory in Norway to facilitate the study of methods for enhancing the
phase separation. Any viable method should address the issue while requiring
little or no energy input, maintain the high thermal output and, to the extent
possible, not sacrifice the simplicity of the DC-TES. In the light of these criteria,
two approaches to achieve rapid phase separation were studied.
The first method involved the use of surfactants as a chemical demulsifier in the
DC-TES. Surfactants were chosen based on solubility parameters and a screening
procedure known as shake-tests. The surfactants effect on the PCM/HTF interfaces
were thoroughly characterized by measuring the interfacial tension using a
pendant drop method. This was in turn connected to the phase separation performance
in the DC-TES, which could be considerably increased by surfactant
addition.
The second method involved the study of drop size on the stability of the
PCM/HTF emulsion. As one of the most important parameters, drop size govern
many aspects of an emulsion, including its stability. However, it also has
huge implications for the motion and heat transfer of HTF drops in a DC-TES.
Manipulating the drop size with the aim of increasing the phase separation must
therefore be done carefully to avoid negatively impacting the thermal performance
of the DC-TES. This was experimentally investigated in three related experimental
campaigns: 1) Investigating the dynamics of drop dispersion with particular
focus on drop sizes in a DC-TES, 2) motion and heat transfer of single HTF drops
descending through PCM, and 3) phase fractions in the DC-TES.
In the first of these three campaigns, drop production from the dynamic disintegration
of the inlet HTF jet was investigated. HTF flow rates were varied, along
with configurations of inlet nozzles. The resulting distribution of drop sizes was
described in terms of appropriate dimensionless numbers. This was subsequently
compared to the phase separation behavior in the DC-TES, where it was found to
have a profound impact.
Investigating drop motion and heat transfer of a single HTF drop was done
using a specialized flow cell and image analysis. Both drop size and ambient PCM
temperature was varied. A review of available models were conducted and compared
to the experimental results. Among the various assumptions adopted by
the models, experimental results highlighted the nature of the interface, i.e rigid
versus mobile, and particle shape as the most significant factors.
Phase fractions in the DC-TES was investigated as the third part of the study
on the effect of drop sizes. This was done using a gamma densitometer setup
to obtain the in situ distribution of phases in the DC-TES. Phase fractions were
measured in both the dispersion region, i.e where heat transfer between the HTF
and PCM occurs, and the densely packed emulsion layer below. Results provided
input for models to extend predictions from single drop to drop dispersions. It
was also found that the rate of continuous phase drainage, as part of the phase
separation process, could be predicted.
In summary, this work thoroughly demonstrates two promising solutions to the
phase separation issue in a DC-TES. The results presented in this thesis advances
the DC-TES concept and establishes a solid basis for up-scaling in the future.