Characterisation of a parabolic trough collector using sheet metal and glass mirror strips

Abstract

A novel type of parabolic trough collector was characterised using a very basic theoretical model. This model looked at an ideal case and provided a basic expectation that was compared to actual measurements. The model showed that greater improvements can be achieved if heat losses to the environment are limited or omitted. This can be achieved by using a glass shield to insulate the receiver in a vacuum to limit the effect wind has and therefore limit convective losses. The experimental characterisation of the PTC consisted of taking six different temperature measurements to better understand the energy balances taking place. Four different configurations were tested, using two different types of concentrator and in each case a receiver that was either unpainted or painted with a semi matte black paint. The different types of concentrator were either stainless steel sheet metal or discretised glass mirror strips, similar to a linear Fresnel collector. Experimental runs were conducted on cloudless days for an hour and 15 minutes. This allowed for three runs to be performed on a single day. Using the theoretical model and comparing it to the experimental data, an efficiency was calculated. This efficiency averaged 14 % when the receiver was unpainted and 13 % when the receiver was painted for the metal sheets. The glass mirror strips had average efficiencies of 54 % and 45 % for an unpainted and painted receiver respectively. The model is very basic and can be improved upon if more variables are taken into consideration, such as convective heat losses. It was also recommended that wind measurements are taken in future tests. A property looked at to evaluate the effectiveness of each type of configuration was the average energy supplied to the thermal heating fluid over the course of an experimental run. For this the averaged values over all the experimental runs conducted for stainless steel sheet metal were 258 W and 332 W for an unpainted and painted pipe respectively. When using the glass mirrors an average energy value of 1049 W was supplied when the pipe was unpainted and an average of 1181 W was gained in the runs conducted after the pipe had been painted. Painting the receiver had little to no effect. The surface temperature of the receiver after painting the pipe was not higher and a slight increase in the energy gained by water was observed. This was explained by inaccuracies during testing as scattered light may have caused an interference on some of the measurements. There were also human inaccuracies in testing which should be omitted in future tests by implementing, for one, a functional tracking system. Future tests should be designed in such a way to completely omit irradiance affecting the thermocouple taking the measurement. Glass mirrors fared far better than the stainless steel sheet metal counterpart. It was recommended that they are used as the concentrator of choice. Higher efficiencies were achieved and in some cases almost four times the energy was supplied to the water in the pipe. This was attributed to a much lower concentrator temperature, on average 11 °C lower than the temperature of the metal sheets, as well as a much better ability to concentrate sunlight onto a single focal point. However, the glass mirror strips were proven to be very fragile and as such, require protection from the elements. While the strips were lighter and caused less of a load during windy conditions, they were susceptible to oscillations from gusty wind. This led to a number of strips breaking and needed to be replaced. By discretising the strips into individual pieces, they had the benefit of only needing to replace the strips that were damaged. This is also true for all future runs. It is still recommended that a tarp be used to protect the glass mirrors. Using glass mirror strips as a concentrator combined LFC technology with PTC technology and a novel PTC design was achieved. The design still required the installation area of a PTC. The novel design was compared to Industrial Solar’s industrial LFC module, LF-11, as it shares many similarities to LFC technology. The peak thermal output of the rig was significantly lower at 346 W/m2 compared to the industrial value of 562 W/m2. However, the noteworthy differences in design and optimisation between the two modules meant the results achieved were comparable. It is expected that better and more comparable results can be realised once the inherent flaws in the design, such as tracking the sun, aperture size and adding a vacuum absorber, are addressed. It is recommended that more research and emphasis is put into this field as an alternative energy power plant for South Africa.