Facility description:

The Plataforma Solar de Almeria gathers the most extended experience in testing volumetric absorbers. More than 20 different materials and geometrical configurations have been qualified since 1986 (Leon, 1996; Tellez, 2002). A versatile test bed allows both the operation of up-to 250 kW solar receivers under real conditions of concentrated solar flux and air mass flow outside the receiver aperture (Tellez, 2003). Some metallic volumetric absorber modules were tested in the Solair Receiver, in the Central Receiver System (CRS) facility at the PSA, Spain. Small solar receivers or prototypes (about 200 kWth ) were tested at the top of the CRS tower. The facility receives the radiation from a
primary field with 91 heliostats of 39.3 m2 of reflecting area and a secondary field in the north with 20 larger heliostats. The nominal average reflectivity of the field is 92%.

The CRS facility has a CCD camera system for measuring the concentrated solar radiation flux. In this system, a Lambertian target is installed over a moving bar. It passes in front of the receiver aperture, parallel to its surface, intercepting the solar irradiation when a flux measurement is required. Two kinds of lambertian targets were used during the tests: a plate of 0.65 m x 0.26 m, area enough to cover the central cups and give the image of the region of interest in only one recorded image, and a Lambertian strip (long and narrow). The image of the concentrated solar focus is given by the superposition of several recorded images. The image of the radiation diffused by the target is obtained and the gray-level value of each pixel is converted to a radiation value by means of the calibration (Ballestrin, 2002). Then, the total power can be integrated and the calculation of the rest of the magnitudes of interest, such as peak irradiance or distribution, is possible.

A calorimetric test-bed for the evaluation of small atmospheric-pressure volumetric receivers is installed at the top of the 43 m metallic tower. The test-bed consists of different sections to conduct the hot air from the absorber to the blower through the water cooled heat exchanger. Every section has the necessary instrumentation to measure the temperature, pressure and flow rate. The test-bed was also implemented with an air — recirculation circuit, to recover the exhaust warm air and to improve the total efficiency.

At the front of the test-bed, the Solair 200 kWth receiver is currently installed (Tellez, 2003). This is a prototype, designed to fit the dimensions of the CRS installation, which is about

0. 9 m aperture diameter. This receiver is based on the same concept that the High Temperature Receivers (HitRec I & II) (Hoffschmidt,1999; Tellez, 2002): a modular receiver of absorber cups held in the rear part by the structure of a double membrane, which serves to re-circulate the air. The Solair receiver improves the material construction of the HitRec, and changes the shape of the cups from hexagons to squares with sides of 125 mm length. The 36 absorber modules are relatively free to move or thermally expand, thanks to a gap between adjacent modules. The gaps are also used to inject the returned air over the absorber entrance at the same time that it cools the stainless steel construction. In addition, they make easier the replacement of the modules.

Tested absorbers:

The Solair receiver concept is not directly linked to one specific absorber material. Any material (ceramic or metallic) can be placed into a cup of the correct dimensions and installed in the receiver. Thus, the collaboration with the German manufacturer of metal catalyst carriers Emitec enabled to test several metallic monoliths of corrugated foil, which were kindly provided by Emitec. The number of modules tested for this work amounts to 31: five square cups and twenty-six cylindrical absorbers (18 cylinders with 8 cm diameter and 8 units with a diameter of 11 cm).

The modules had different lengths, ranging from 20 mm to 58 mm in the case of the cylindrical absorber modules. Moreover several cell densities were tested, varying between 400 and 800 cells (or channels) per square inch (cpsi). All the tests for the different absorber modules were carried out during two test campaigns at the PSA: the first in July 2003 and the second in the winter of 2003-04. The first test campaign was aimed to measure the air outlet temperature distribution. Thus, the metallic square cups of 125×125 mm2 were instrumented with 7 thermocouples type K, which were symmetrically placed at three different radial distances in a cross section located at 2 cm from the absorber exit.

The set of 8-cm diameter cylindrical absorber modules were adapted to the ceramic square cups with an alumina plate adaptor. Due to the smaller area of the absorbers, only 5 sensors were placed at the absorber exit. These thermocouples were located also symetrically at two different radial distances.

In the second campaign, besides the thermocouples placed at the absorber exit, some additional thermocouples were inserted inside the absorber. The positions for the thermocouples inside the absorber were planned regarding the results of the model of Hoffschmidt (1996). These results state that the channels in the absorber centre transfer the heat radially to the edge of the absorber module, resulting in different axial temperature distributions for a channel in the absorber axis or near the absorber edge. In order to measure these axial temperature differences, in the 11-cm diameter cylindrical modules, 4 thermocouples were inserted through central channels at different downstream distances. Other 3 thermocouples were put near the edge.

The last two square metallic absorber cups tested were instrumented during manufacturing (see fig.1) with 5 inner thermocouples about the absorber centre at 1, 3, 10, 30 and 50 mm from the entrance, respectively (see thermocouples 1-5 at fig.1). Other 3 thermocouples were installed at a distance of 30 mm from the central axis, at 50 mm, 30 mm and 10 mm from the entrance (see thermocouples 6-8). Behind the absorber outlet section, other 7 themocouples were installed: one at the centre, two at 30 mm from the centre at both sides and four at 60 mm in the diagonal lines. One of these cups had a cell density of 600 cpsi and the other 500 cpsi. The absorber length inside the metallic square cup was 7 cm.

Planning of the tests:

The experimental goal was to perform a parametric study of the absorber temperature response with the variation of the most relevant system inputs: the flux of the concentrated radiation onto the absorbers and the cooling air mass flow rate. The usual test procedure was to begin the test with a certain number of heliostats and the maximum mass flow rate (blower power set to 100%). After approximately one hour, when the whole receiver had heated up and achieved a steady temperature, the incoming radiation was measured for those conditions. Then, the mass flow rate was reduced by decreasing the blower power by 5% or 10%. About twenty minutes later, when the cups outlet air temperatures had
achieved a new steady value (provided that no clouds disturbed the system), a new flux measurement was taken. Furthermore, the mass flow rate was lowered again, repeating the process for several values of the blower power. When the weather conditions were good and no clouds obliged to wait until having stable direct radiation and high temperatures again, five or six different values of the air flow rate were used (reaching the 70% or 75% of the blower power). If there was still some time before ending the working day, the mass flow rate was increased up to a value already tested before, in order to check repeatability. Some other specific tests were also made, like changing the air return rate or decreasing radiation and mass flow rate at the same time.

The data acquisition system (DAS) of the calorimetric test-bed with the Solair receiver has 138 signal inputs. Due to large amount of data that produces one single record, these data were recorded every five seconds. The acquisition system was implemented with the sensors installed inside the cups with metallic absorbers, but these signals were recorded every second, in order to appreciate any unstable behaviour, even with a very small frequency. Another temperature was recorded in the implemented DAS: the ambient temperature just under the receiver, given by a thermocouple out of the focus of the incoming radiation. It is usually assumed in the literature that the air temperature at the absorber entrance coincides with the ambient temperature, disregarding the possible heating up of the air around the receiver due to radiation losses.