Design options for separate reactors

An important design specification for a TCM system is the required thermal power input, which is related to the amount of material in the reactor and how fast the solar heat can be fed to the material. Typically, this process is complicated by the low thermal bed conductivity of the TCM powder, which is due to the low conductivity of the gas in the pores between the particles. Also the water vapour transport out of the layer of TCM powder is critical, and insufficient vapour transport out of the material may result in unwanted melting. One can try to improve the heat and vapour transfer simultaneously by stirring, thereby speeding up the conversion. Three design cases have been evaluated for dehydration reactor volume in the following paragraphs, under the assumptions that the reactor power required is 3 kW (see Fig 4) and that heat transport is the limiting factor.

4.1 TCM fluidised bed reactor

A fluidised bed reactor (see Fig. 5a) creates very good mixing of the powder, thereby improving strongly the vapour transport and the heat transfer. Fig. 5b shows that for small particles, the fluidisation velocity is small, limiting the power required for the fluidisation, while the heat transfer to the embedded heat exchanger is strongly increased. However, the fluidisation may cause breakup of TCM particles that are fragile already due to the cracks caused by repetitive hydration and dehydration. The resulting fine dust may be blown out of the fluidised bed reactor, thereby reducing the amount of active material. Also, this very fine powder cannot be fluidised properly.

For the case of dehydration of the TCM in a 3 kW reactor, assuming 200 micron particles and an effective heat transfer of about 350 W/m2K to the immersed plate heat exchanger (consisting of 11 parallel plates of 7.5 cm x 19 cm each), this would lead to a reactor fluidisation section (excluding the freeboard and the gas distribution section) of about 20 cm in length and 11 cm diameter, in which about 40% of the open volume would be filled with TCM, while about 30% of the reactor volume is taken up by the heat exchanger volume. Due to the low fluidisation velocity (resulting from the small particle size), the power required for fluidisation is in the order of a few Watt. This seems promisingly low, provided that the pressure drop in the porous support and other parts of the reactor can also be kept small.