Thermal behaviour of granular materials in directly irradiated fluidized beds

The production and storage of renewable energy is a top challenge for mankind. Extensive R&D is in progress to exploit the huge amount of solar energy falling on Earth (~100,000 TW). Thanks to the easy integration with Thermal Energy Storage (TES), Concentrating Solar Power (CSP) systems stem out as one the most promising technology to take advantage of the solar energy. TES allows to separate the process of usage and absorption of solar energy: in this way, the energy coming from the sun can be exploited at will, even during the night or in cloudy sky conditions. The current benchmark in the CSP technology is represented by the solar tower receiver with molten salts working as heat transfer fluid and sensible heat storage medium. The main drawback of this system is the relatively low working temperature of the molten salts (about 565 °C), which affects the efficiency of the subsequent Rankine cycle for energy production. Particle receivers in CSP systems are gaining ever increasing interest, as the solid particles can simultaneously act as receiver, heat transfer fluid and heat storage medium. Dense gas solid suspensions can work at higher temperature with respect to molten salts, even between 1000–1500 °C, without any technological problem. Fluidized beds (FBs) as dense gas solid suspensions are considered as affordable and reliable systems for CSP. In solar FB, the interaction between the incident radiative flux and the FB can occur in an indirect or direct way. In the indirect heating, the concentrated radiation is focused onto a cavity or an exposed surface, and heat is eventually transferred to the solid suspension by convection/conduction. In the direct heating, the concentrated radiation enters the reactor by the use of optically accessible windows located in the ceiling of the receiver, or through the use of transparent walls. The direct heating configuration permits higher operating temperatures as the high concentrated solar radiation directly heats the dense gas solid suspension, but care must be paid to avoid the bed surface overheating. In this work, different materials were investigated to scrutinize their potential use as solid suspension in directly irradiated FB reactors. The experimental apparatus consists of a FB reactor with both the internal bed diameter and height equal to nearly 0.1 m, so that the aspect ratio (internal diameter-height ratio) approaches the unit value. The windbox section, below the reactor zone, is 0.15 m high. The upper part of the FB reactor is connected to a conically-shaped section, 0.4 m high, which represents the freeboard and hosts, at is extremity, a 4 mm transparent window resistant to high temperature, essential to let the solar simulated radiation enter. The conical shape of the freeboard section is required to both not hinder the pathway of the solar simulated radiation and to further protect the transparent window against particles impact, thanks to the gas slowdown due to the cross-section increase. Gas exit is provided at half the conical section through four 1 inch pipes. The FB reactor surface is exposed to a 12 kWel simulated solar furnace made by an array of three short-arc Xe-lamps coupled with elliptical reflectors. The peak flux produced on the FB surface is of nearly 3000 kW m–2, while the total power irradiated over the whole FB surface is of about 3 kWth when all the lamps are turned on. On the other side, when only two lamps are used, the value of peak flux and total power becomes 2000 kW m–2 and 2 kWth, respectively. The reactor is also surrounded by two semi cylindrical radiant heaters in ceramic fiber, which are used to both heat and insulate the reaction chamber. An electronic mass flow controller is used to supply the air required for the reactor operation. Concerning the diagnostic tools, five K-type thermocouples are located inside the FB to measure the temperature at different locations. The experimental tests were aimed at studying the dynamics of the directly irradiated FB reactor with specific reference to temperature distribution at the surface and in the bulk of the bed as a function of the inlet gas velocity. The tests were performed by powering only two of the three Xe-lamps because the temperature level reached with this configuration were already sufficiently high for the exploitation of thermodynamic cycles for energy production. In each test, the bed was initially charged with an amount of material so that the static bed height was equal to 7.8 cm. To speed up the initial warm up of the reactor, the system was heated by using the radiant heaters up to a temperature of 400 °C while fluidizing with an inlet gas velocity of 10 cm s-1. Then, the radiant heaters were turned off and simultaneously the solar simulator was powered on. It was then waited until all the temperature values read by the K-type thermocouples become stationary. Once the steady state condition was obtained, the inlet gas velocity was reduced by 1 cm s–1 to eventually reach a new stable operation condition. Typically, a time of 20 min was always sufficient to achieve a new steady state operation condition. Following this method, the inlet gas velocity was progressively reduced to the lowest possible stable value according to the fluidization behaviour of the investigated material. During the tests, the temperature data were continuously acquired at 1 Hz using a purposely developed script under Labview environment. Data of steady-state bed temperature as a function of the inlet gas velocity were collected for the different materials and eventually used to compare the material performance and to estimate an efficiency of absorption of the solar energy. More into detail, the data were used to solve an energy balance model on the FB reactor whose solution returns the absorption thermal efficiency, as a function of the inlet gas velocity, for all the investigated material. Experimental results highlighted that steady-state temperatures tend to decrease when the inlet gas velocity was increased. This behaviour was attributed to both an increase in the thermal losses due to convective flow of the fluidizing gas and to the reduction of absorption capacity of the bed surface, as probably clouds of particles ejected by bubble bursting could hinder the penetration of the concentrated radiation to the FB. However, by reducing the inlet gas velocity, temperature gradients inside the fluidized bed are increased due to a reduction of the axial thermal diffusivity in the FB reactor. By solving the energy balance equation on the FB reactor under steady state conditions, the overall absorption coefficient for the investigated materials was estimated. It was then possible to rank the materials according to their efficiency in absorbing the incident solar energy. It was also observed that, as for the steady state temperature, the absorption coefficient increased when the fluidization velocity was reduced. However, to properly choose the best material to be used as granular material in directly irradiated fluidized beds, it is important to take into consideration also the lowest fluidization velocity at which the material can work to withstand the high concentrated solar flux. Indeed, gas pumping and compression represent one of the major costs in FB reactors, thus a lower inlet gas velocity brings to lower operational costs.