- Email: [email protected]
A new solar concentrating system: Description, characterization and applications
Available online at www.sciencedirect.com
Solar Energy 85 (2011) 1000–1006 www.elsevier.com/locate/solener
A new solar concentrating system: Description, characterization and applications J. Llorente a,⇑, J. Ballestrı´n b, A.J. Va´zquez a a
Corrosion and Protection Department, National Center for Metallurgical Research (CENIM-CSIC), Avda. Gregorio del Amo 8, 28040 Madrid, Spain b CIEMAT-Plataforma Solar de Almerı´a, Aptdo. 22, E-04200 Tabernas, Almerı´a, Spain Received 16 October 2009; received in revised form 18 February 2011; accepted 22 February 2011
Communicated by: Associate Editor L. Vant-Hull
Abstract Solar concentrating systems are usually very expensive and require a large space for their installation. This article presents a new solar concentrating device which is low-cost, small-scale, and has very good features for materials treatment. It consists of two sets of mirrors that reflect solar radiation in two steps with a beam array similar to a Fresnel lens. The power density was measured with Gardon-type radiometers. The results are in good agreement with previous work. The system has a nominal power of 2.5 kW, a measured concentration factor of 1040, and a measured focal diameter of 20 mm (90% of power level). Ó 2011 Elsevier Ltd. All rights reserved. Keywords: Concentrated solar energy; Solar flux measurement; Solar concentrator; Solar processing; Heat treatments
1. Introduction Since the late 70s, concentrated solar energy has been used successfully in materials processing, but only in research. During these years, many processes, such as steel surface thermal treatments (Va´zquez et al., 1991; Rodrı´guez, 1997), synthesis of wear and corrosion resistant metal coatings (Sierra and Va´zquez, 2006; Ferriere et al., 2006) and modification of ceramic materials have been developed (Costa Oliveira et al., 2005). Today, the usefulness of CSE in all those processes has been demonstrated and first pre-industrial applications are being tested (Can˜adas et al., 2006). (See Flamant et al., 1999 for a brief review.) In solar materials processing, the most commonly used technology is the “solar furnace”. It consists of one or several heliostats which reflect solar radiation onto a fixed parabolic mirror which in turn concentrates the rays in its focus in a high power density beam. There are some varia⇑ Corresponding author.
E-mail address: [email protected] (J. Llorente). 0038-092X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.solener.2011.02.018
tions on this, such as secondary concentrators, and off-axis designs. For examples of these systems, see (Lewandowski et al., 1991; Riskiev and Suleimanov, 1991; Neumann and Groer, 1996; Guesdon et al., 2006). All of them require a large site for the heliostats and the parabolic mirror housing, and they are also high-cost systems which are very difficult to operate and maintain. There are other installations that are comparatively cheaper, smaller and easier to use. A small parabolic dish, for example, can produce very high concentration. But the main disadvantage of this system for materials treatment is the position of the focus at the top, making it necessary for samples to be set face down. This is unsuitable for treatments involving melting of the material, for example. A more suitable device is the refractive Fresnel lens, which consists of a 1 m2 multifaceted polymeric disk which refracts solar radiation and concentrates the rays on its focus. Concentration factors and temperatures achieved by this system are comparable or higher than those achieved by larger, more expensive facilities (Sierra and Va´zquez, 2005; Ferriere et al., 2004). It has been used
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006
1001
successfully in materials processing since the mid 90s for applications such as thermal treatment, TiN coatings on Ti6Al4V alloy (Sa´nchez Olı´as et al., 1999) or NiAl coatings on steel by self-propagated high-temperature synthesis (SHS) Sierra and Va´zquez, 2005. This paper presents an alternative small-scale, low-cost facility for concentrating solar energy.
2. Experimental 2.1. Description of the new system The new solar concentrating technology is called “Double Reflection Fresnel Lens” (DRFL) (Fig. 1). It is comprised of three main parts: Optics, consisting of all the mirrors that reflect and concentrate the solar rays in the focus. Electronics, including the solar-tracking timer. Mechanics, consisting of the equatorial mounting and the structure that enable the solar tracking of optics by means of the electronics. The optics consist of 864 mirrors (approx.15 12 cm) in nine concentric rings on a circular (Ø = 3.5 m) plane normal to the solar radiation. There are two sets of mirrors. The first set is fixed at a 45° angle from the optical axis with the reflected rays normal to it, and a second set of parabolic mirrors (nominal 8 concentration factor, i.e. the power density in the focus of each mirror is eight times the incident power density) reflects these rays, concentrating them on the focus. The mirrors from the second set have different focal length depending on their position but all of them have an on-axis parabolic shape. Their inclination is varied manually one by one depending on their distance from the optical axis to concentrate all the rays in the center of a ring delimiting the focal plane (Fig. 2). This ring has several holes that can be used to attach acces-
Fig. 1. A general view of the “Double Reflection Fresnel Lens”.
Fig. 2. DRFL beam array.
sories for the experiments or maintenance, such as focusing or system orientation (Fig. 3). The optics are connected to the ring by a tube frame and both are properly counterbalanced to ensure system stability during solar tracking. The base is an equatorial mounting which makes one-axis solar tracking possible. This mounting is oriented by the polar axis at the same inclination angle as the site latitude (40.4° N for Madrid). A curved screw pulls the structure up or down depending on the solar elevation, which varies with the day of the year, and is highest in summer and lowest in winter. This height is regulated by hand once at the beginning of the day and remains almost constant during the day. The entire system is moved by a wheel connected to a small stepping motor controlled by a timer, which regulates the interval and duration of movement. This movement tracks the sun, keeping the position of the focus fixed in the center of the ring. Some optical analysis can be performed based on Fig. 4. The real dimensions of all the mirrors and the angles they
Fig. 3. Detailed view of the focal ring, with holes for attaching measurement or experimental accessories.
1002
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006
The power in the different stages of reflections is the following: P r ¼ I r S1 ¼ RI 0 cos 45 S1
ð3Þ
P f ¼ I f S2 ¼ R2 I 0 cos 45 sin bS2
ð4Þ
where R is the mirror reflectivity, I0 is the direct solar irradiance, Ir is the irradiance reflected by the mirrors in the first set, If is the irradiance reflected by the mirrors in the second set, Pr is the power reflected by the mirrors in the first set, Pf is the power reflected by the mirrors in the second set, S1 is the mirror surface of the first set (S1mirror number of mirrors), S2 is the mirror surface of the second set (S2mirror number of mirrors). Table 3 gives Pr/R I0 and Pf/R2 I0 per ring. Adding up the contributions from each row, Pf can be expressed as:
Fig. 4. Geometric drawing for theoretical analysis.
P f ¼ 39445R2 I 0 ðI 0 in W=cm2 Þ form with the optical axis were measured. Table 1 gives the number of mirrors per ring in the first set, where ring 1 is the closest to the center, as well as the surface of each mirror and the effective collecting surface in each ring, taking the cosine factor into account as follows:
So, for a typical mirror reflectivity R = 0.8 with a direct solar irradiance of 0.1 W/cm2, the total power in the focus of the concentrator is 2.5 kW. The concentration factor can also be calculated. The irradiance in each stage can be expressed as:
S1effective; ring ¼ ðnumber of mirrors per ringÞ ðsurface of each mirrorÞ cos 45
I r ¼ RI 0 cos 45
ð6Þ
I f ¼ C mirrors2ndgroup R2 I 0 cos 45 sin b exp½2ðr=2rÞ2
ð7Þ
ð1Þ
From the data in Table 1, adding all the contributions from the third row, the effective collecting surface is 5.42 m2, so for a typical irradiance of 1000 W/m2, the collected power is 5.42 kW. However, as the rays are reflected twice and the incidence is not normal, mirror reflectivity losses and the different angles have to be considered. Table 2 lists the geometric data for the second set of mirrors (see Fig. 5). In this case, the equation for S2effective per ring is:
Table 4 lists Ir and If at the focal point (r = 0), assuming that the mirrors in the second set have a concentration factor of 8. The contribution of each ring to the concentration factor must be corrected for the angle formed by the optical axis and the corresponding ring (1–9), which has been labeled the h angle. Adding up the contributions from all the rings, the irradiance at the focus as a function of mirror reflectivity and the direct solar irradiance can be expressed as:
S2effective per ring ¼ ðnumber of mirrors per ringÞ ðsurface of each mirrorÞ sin b
ð5Þ
I focal ¼ 1910R2 I 0 ðwith I 0 in W=cm2 Þ
ð2Þ
ð8Þ
Table 1 Effective surface of the first set of mirrors. Ring
1
2
3
4
5
6
7
8
9
Number of mirrors S1mirror (cm2) S1effective per ring (cm2)
16 186 2104
24 186 3156
32 186 4209
40 186 5261
48 186 6313
56 172.5 6831
64 172.5 7806
72 172.5 8782
80 172.5 9759
Table 2 Effective surface of the second set of mirrors. Ring
1
2
3
4
5
6
7
8
9
Number of mirrors b angle S2mirror (cm2) S2effective per ring (cm2)
16 50 168 2059
24 51.875 168 3172
32 53.75 168 4335
40 55.625 156 5149
48 57.5 150 6072
56 59.375 150 7228
64 61.25 144 8080
72 63.125 144 9248
80 65 144 10,441
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006
1003
Table 4 Contribution by each row to irradiance in the different stages. Ring
1
2
3
4
5
6
7
8
9
Ir/(R I0) h If/(R2 I0) Ifocal/(R2 I0)
11 8 69 68
17 11 107 105
23 14 146 142
28 17 187 179
34 20 229 215
40 23 273 251
45 26 317 285
51 29 363 317
57 32 410 348
Fig. 5. Geometrical drawing of irradiance calculation in the focal plane.
For a reflectance of 0.8, the concentration factor is 1222 (i.e. the power density in the focus of the concentrator is 1222 times the direct solar irradiance), and for a reflectance of 0.7, the concentration factor is 936. 2.2. Measuring the system Gardon-type radiometers were used to map power density (Figs. 6a and 6b). These sensors are basically a differential thermocouple that measures the temperature difference between the center and the circumference of a thin circular foil disk (Gardon, 1953). The signal sent by the sensor is simply measured with a voltmeter (accuracy ±0.1 mv). The direct solar irradiance was measured continuously throughout testing with a Kipp and Zonen CH1 pyrheliometer (accuracy ±0.5%) and PT100 temperature sensor to correct the irradiance measurement. The Vatell Corp. is the only company currently manufacturing this kind of radiometer. They come with two types of black coating depending on the flux measurement range, ZynolyteÒ below 3500 kW/m2, and colloidal graphite above that. In this work, both types were used, taking into account the different spectral emissivity of the two coatings. However, the spectral distribution used in calibration is different from the solar spectrum, overestimating solar flux measured as a result. Therefore, when measuring concentrated solar radiation with such sensors a correction must be made. For radiometers with the ZynolyteÒ coating, the calibration constant correction factor is 0.965 and for colloidal graphite coating, it is 0.782 (Ballestrin et al., 2003). The accuracy quoted by Vatell for the radiometers was ±3%. The experimental device shown in Fig. 7 was designed to place the radiometers and move them around the focus. An elevator type scheme was chosen to provide movement on three axes. Movement along the optical axis (Z axis) was
Fig. 6a. Commercial Gardon-type radiometer.
Fig. 6b. Detail of the black radiometer coating.
electronically controlled, and movement on the XY plane was done manually. This device was attached to one of the holes in the focal plane delimiting ring in such a way that the positions of x and y in the [30, +30] interval could be measured with zero as the focal point. A polished stainless steel tube with holes in it to facilitate heat dissipation was connected to this device. The radiometer was
Table 3 Power contribution of each row of mirrors. Ring
1
2
3
4
5
6
7
8
9
Pr/(R I0) Pf/(R2 I0)
2104 1456
3156 2242
4209 3066
5261 3641
6313 4294
6831 5111
7806 5713
8782 6539
9758 7383
1004
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006
Fig. 9. Qualitative tests of two different specimens. (A) A 20 mm-dia. hole in a 0.5 mm-thick steel sheet after 40 s. (B) A 60 mm hole in a 1 mm-thick aluminum sheet after 10 s.
inserted in the tube, with cooling tubes shielded from direct radiation and insulated from heat with alumina (low thermal conductivity) and highly reflective aluminum foil (Fig. 8). 3. Results and discussion When all the mirrors had been properly focused, some qualitative tests were made. The first measured the temperature on a steel sheet used for focusing the mirrors with a Type K thermocouple. Temperatures up to 1500 K were recorded under good solar irradiance conditions. The second qualitative test was melting two different metals. A 0.5 mm-thick steel sheet was melted in 30–40 s, making a 20 mm-dia. hole in the center. A 1 mm-thick aluminum sheet was melted in just 10 s making a 60 mm-hole in the center. Both tests showed good beam symmetry, as seen in Fig. 9. As mentioned above, two types of radiometers were used for measuring solar power density, one with a colloidal graphite coating labeled 8170 and the other with a ZynolyteÒ coating labeled 8168. Fig. 10 shows the varia-
Fig. 8. Experimental setup used for the radiometer housing.
Power density (kW/m2)
Fig. 7. Radiometer housing device designed.
φpow=(1040
1250 1200 1150 1100 1050 1000 950 900 850 800 750 700 650 0.70
0.75
0.80
dir
0.85
0.90
0.95
1.00
2
Direct solar irradiance (kW/m ) Fig. 10. Direct irradiance vs. power density measured with two different radiometers. The concentration factor is 1040.
tion in solar power density in the focus (/pow is the power density and Idir is the direct solar irradiance). Experimental data from both radiometers with and without correction are shown. Only if the correction factor (Section 2.2) for the spectral emissivity of the two coatings is taken into account can the data from both radiometers be compared and a linear fit performed. These results show that the measured concentration factor is 1040 ± 60, which is quite near the one calculated with Eq. (8), giving an idea of the focusing quality, although it could still be slightly improved. Both types of radiometer were measured several times at nearly 300 different points all around the focal point, including the three axes, for a complete description of the power density distribution in the focus. When x had been set, a scan was made of the Z axis. When the scan was finished, x was reset and the Z axis was scanned again, and so on. The data for the Z = 0 focal plane can be fitted to a Gaussian curve (Fig. 11). Due to the good beam symmetry, this fit can be extrapolated to the whole space, and the power density in the focal plane can be expressed as:
/pow
r 2 ¼ C I dir exp 2 2r
ð9Þ
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006 Gaussian fit Experimental data
Normalized power density
1.0 0.8
C90
0.6 0.4 0.2 0.0 -80
-60
-40
-20
0
20
40
60
80
X axis (mm) Fig. 11. Normalized power density distribution in the focal plane and its Gaussian fit.
Relative power density
1.0 0.9 0.8 0.7 0.6 0.5 0.4 -100 -80
-60
-40
-20
0
20
40
60
80
Distance to the focal point along the optical axis (mm) Fig. 12. Power density variation along the optical axis. Positive axis faces the sun.
where /pow is the power density, C is the concentration factor of 1040, Idir is the direct solar irradiance, r is the distance to the focal point in mm and r is the standard deviation, which is (20 ± 2) mm. Defining the focal spot as the region where the power density is at least 90% of the maximum, it may be seen from the graphs that the diameter of the focal spot of this equipment is 20 mm. Variation of power density along the optical axis is shown in Fig. 12. It may be seen that there is nearly no variation in an interval of ±20 mm from the focal point. This
1005
feature makes the concentrator beam very suitable for application to materials treatments, including specimens larger than those previously treated with other concentrating technologies. The new concentrating system and the systems described in the introduction are compared in Table 5. As the main application of the new device is materials treatment, only the focus diameter and the power density need to be considered. The DRFL has a larger focus diameter than the Fresnel lens, although the power density is substantially lower. Nevertheless, the temperature achieved with the DRFL is high enough to conduct a wide variety of materials treatments. According to the table, the power density and focus diameter are considerably lower than other solar concentrators (central receivers, parabolic mirrors or solar furnaces), but the scale of the DRFL is a good advantage which should be taken into account. Another advantage over parabolic dishes is that due to the position of the focus, samples can be positioned almost face up. This new solar device is a good alternative for solar materials treatment mainly due to its small-scale and the focus diameter. The concentrator presented is only a prototype and can be improved in several respects. As mentioned above, each mirror is positioned manually, so how accurately this is done strongly affects the results, particularly the concentration factor. As mirrors get dirty, the concentration factor decreases, but when they are cleaned, the mirrors may be moved slightly out of position, losing focusing accuracy. A compromise must therefore be reached between cleaning and focusing. Two months between consecutive cleanings is good enough to keep focusing accurate and a good concentration factor. In future designs, the diameter of the focus could be reduced by changing the mobile mirror quality to increase the power density. A much better concentration factor would be achieved this way, although the cost of the equipment would increase. The optics could also be enlarged to increase system power, but one of the main advantages of the DRFL, which is its small-scale, would then be lost. 4. Conclusions A new low-cost and easy-to-use, small-scale solar concentrator has been completely described and characterized.
Table 5 Comparison of different solar concentrating systems. Kind of installation
Examples
Focus diameter (mm)
Power density (kW/m2)
Central tower and field of heliostats Parabolic mirror Solar furnace Fresnel lens Double Reflection Fresnel Lens
Almerı´a (PSA-CIEMAT), Themis (IMP-CNRS), Alburquerque, (Sandia National laboratories) Almerı´a (PSA-CIEMAT), Schengen, Suiza (PSI) Almerı´a (PSA-CIEMAT), Odeillo (IMP-CNRS). PSI Madrid (CENIM-CSIC), Ciudad Real (ETSII-UCLM), Lisboa (IST) Madrid (CENIM-CSIC)
1.5–3
700–3000
1–120 5–230 8 20
2000–4000 2500–17,000 2600 1000
1006
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006
Flux mapping was done with Gardon-type radiometers and an elevating device to move the sensor around the focus. Two types of Gardon-type radiometers (ZynolyteÒ and colloidal graphite) were used showing good agreement with other studies on solar flux measurement. The concentration factor was 1040, with a 20 mm focal diameter at 90% power. The power density achievable and the focal diameter make this equipment suitable for materials treatments. Acknowledgements This work was funded by the Spanish Ministry of Education and Science Under Project ENE2004-07502, “High Temperature Concentrated Solar Energy Applications”. The new concentrating system described in this work was designed, patented and manufactured by Cruz and Bomant S. L. References Ballestrin, J., Ulmer, S., Morales, A., Barnes, A., Langley, L.W., Rodriguez, M., 2003. Systematic error in the measurement of very high solar irradiance. Solar Energy Materials & Solar Cells 80, 375– 381. Can˜adas, I., Martinez, D., Rodrı´guez, J., Fernandez-Gonzalez, B.J., Va´zquez Vaamonde, A.J., 2006. Procesamiento en lechos fluidizados calentados con energı´a solar concentrada. In: Proceedings of XII Congresso Iberico e VII Ibero Americano de Energia Solar Lisbon, Portugal. Costa Oliveira, F.A., Shohoji, N., Cruz Ferna´ndez, J., Guerra Rosa, L., 2005. Solar sintering of cordierite-based ceramics at low temperatures. Solar Energy 78, 351–361. Ferriere, A., P Rodriguez, G., Sobrino, J.A., 2004. Flux distribution delivered by a fresnel lens used for concentrating solar energy. Journal of Solar Energy Engineering – Transactions of the ASME 126 (1), 654– 660.
Ferriere, A., Sa´nchez Bautista, C., Rodrı´guez, G.P., Va´zquez, A.J., 2006. Corrosion resistance of stainless steel coatings elaborated by solar cladding process. Solar Energy 80 (109), 1338–1343. Flamant, G., Ferriere, A., Laplaze, D., Monty, C., 1999. Solar processing of materials: opportunities and new frontiers. Solar Energy 66 (2), 117–132. Gardon, R., 1953. An instrument for the direct measurement of intense thermal radiation. Review of Scientific Instruments 24 (5), 366–370. Guesdon, C., Alxneit, I., Tschudi, H.R., Wuillemin, D., Petrasch, J., Brunner, Y., Winkel, L., Sturzenegger, M., 2006. PSI´s 1 kW imaging furnace – A tool for high – temperature chemical reactivity studies. Solar Energy 80 (10), 1344–1348. ´ Gallagher, J., Winston, R., Sagie, D., Lewandowski, A., Bingham, C., O 1991. Performance characterization of the SERI high-flux solar furnace. Solar Energy Materials 24, 550–563. Neumann, A., Groer, U., 1996. Experimenting with concentrated sunlight using the DLR solar furnace. Solar Energy 58, 181–190. Riskiev, T.T., Suleimanov, S.K.H., 1991. Double mirror polyheliostat solar furnace of 1000 kW thermal power. Solar Energy Materials 24, 625–632. Rodrı´guez, G.P., de Damborenea, J.J., Va´zquez, A.J., 1997. Surface hardening of steel in a solar furnace. Surface and Coatings Technology 92, 165–170. Sa´nchez Olı´as, J., Garcı´a, I., Va´zquez, A.J., 1999. Synthesis of TiN with solar energy concentrated by a Fresnel lens. Materials Letters 38, 379– 385. Sierra, C., Va´zquez, A.J., 2005. High solar energy concentration with a Fresnel lens. Journal of Materials Science 40, 1339–1343. Sierra, C., Va´zquez, A.J., 2005. NiAl coatings on carbon steel by selfpropagating high-temperature synthesis assisted with concentrated solar energy: mass influence on adherence and porosity. Solar Energy Materials and Solar Cells 86, 33–42. Sierra, C., Va´zquez, A.J., 2006. Dry sliding wear behaviour of nickel aluminides coatings produced by self-propagating high-temperature synthesis. Intermetallics 14, 848–852. Va´zquez, A.J., Rodrı´guez, G.P., De Damborenea, J., 1991. Surface treatment of steels by solar energy. Solar Energy Materials 24, 751– 759.
Solar Energy 85 (2011) 1000–1006 www.elsevier.com/locate/solener
A new solar concentrating system: Description, characterization and applications J. Llorente a,⇑, J. Ballestrı´n b, A.J. Va´zquez a a
Corrosion and Protection Department, National Center for Metallurgical Research (CENIM-CSIC), Avda. Gregorio del Amo 8, 28040 Madrid, Spain b CIEMAT-Plataforma Solar de Almerı´a, Aptdo. 22, E-04200 Tabernas, Almerı´a, Spain Received 16 October 2009; received in revised form 18 February 2011; accepted 22 February 2011
Communicated by: Associate Editor L. Vant-Hull
Abstract Solar concentrating systems are usually very expensive and require a large space for their installation. This article presents a new solar concentrating device which is low-cost, small-scale, and has very good features for materials treatment. It consists of two sets of mirrors that reflect solar radiation in two steps with a beam array similar to a Fresnel lens. The power density was measured with Gardon-type radiometers. The results are in good agreement with previous work. The system has a nominal power of 2.5 kW, a measured concentration factor of 1040, and a measured focal diameter of 20 mm (90% of power level). Ó 2011 Elsevier Ltd. All rights reserved. Keywords: Concentrated solar energy; Solar flux measurement; Solar concentrator; Solar processing; Heat treatments
1. Introduction Since the late 70s, concentrated solar energy has been used successfully in materials processing, but only in research. During these years, many processes, such as steel surface thermal treatments (Va´zquez et al., 1991; Rodrı´guez, 1997), synthesis of wear and corrosion resistant metal coatings (Sierra and Va´zquez, 2006; Ferriere et al., 2006) and modification of ceramic materials have been developed (Costa Oliveira et al., 2005). Today, the usefulness of CSE in all those processes has been demonstrated and first pre-industrial applications are being tested (Can˜adas et al., 2006). (See Flamant et al., 1999 for a brief review.) In solar materials processing, the most commonly used technology is the “solar furnace”. It consists of one or several heliostats which reflect solar radiation onto a fixed parabolic mirror which in turn concentrates the rays in its focus in a high power density beam. There are some varia⇑ Corresponding author.
E-mail address: [email protected] (J. Llorente). 0038-092X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.solener.2011.02.018
tions on this, such as secondary concentrators, and off-axis designs. For examples of these systems, see (Lewandowski et al., 1991; Riskiev and Suleimanov, 1991; Neumann and Groer, 1996; Guesdon et al., 2006). All of them require a large site for the heliostats and the parabolic mirror housing, and they are also high-cost systems which are very difficult to operate and maintain. There are other installations that are comparatively cheaper, smaller and easier to use. A small parabolic dish, for example, can produce very high concentration. But the main disadvantage of this system for materials treatment is the position of the focus at the top, making it necessary for samples to be set face down. This is unsuitable for treatments involving melting of the material, for example. A more suitable device is the refractive Fresnel lens, which consists of a 1 m2 multifaceted polymeric disk which refracts solar radiation and concentrates the rays on its focus. Concentration factors and temperatures achieved by this system are comparable or higher than those achieved by larger, more expensive facilities (Sierra and Va´zquez, 2005; Ferriere et al., 2004). It has been used
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006
1001
successfully in materials processing since the mid 90s for applications such as thermal treatment, TiN coatings on Ti6Al4V alloy (Sa´nchez Olı´as et al., 1999) or NiAl coatings on steel by self-propagated high-temperature synthesis (SHS) Sierra and Va´zquez, 2005. This paper presents an alternative small-scale, low-cost facility for concentrating solar energy.
2. Experimental 2.1. Description of the new system The new solar concentrating technology is called “Double Reflection Fresnel Lens” (DRFL) (Fig. 1). It is comprised of three main parts: Optics, consisting of all the mirrors that reflect and concentrate the solar rays in the focus. Electronics, including the solar-tracking timer. Mechanics, consisting of the equatorial mounting and the structure that enable the solar tracking of optics by means of the electronics. The optics consist of 864 mirrors (approx.15 12 cm) in nine concentric rings on a circular (Ø = 3.5 m) plane normal to the solar radiation. There are two sets of mirrors. The first set is fixed at a 45° angle from the optical axis with the reflected rays normal to it, and a second set of parabolic mirrors (nominal 8 concentration factor, i.e. the power density in the focus of each mirror is eight times the incident power density) reflects these rays, concentrating them on the focus. The mirrors from the second set have different focal length depending on their position but all of them have an on-axis parabolic shape. Their inclination is varied manually one by one depending on their distance from the optical axis to concentrate all the rays in the center of a ring delimiting the focal plane (Fig. 2). This ring has several holes that can be used to attach acces-
Fig. 1. A general view of the “Double Reflection Fresnel Lens”.
Fig. 2. DRFL beam array.
sories for the experiments or maintenance, such as focusing or system orientation (Fig. 3). The optics are connected to the ring by a tube frame and both are properly counterbalanced to ensure system stability during solar tracking. The base is an equatorial mounting which makes one-axis solar tracking possible. This mounting is oriented by the polar axis at the same inclination angle as the site latitude (40.4° N for Madrid). A curved screw pulls the structure up or down depending on the solar elevation, which varies with the day of the year, and is highest in summer and lowest in winter. This height is regulated by hand once at the beginning of the day and remains almost constant during the day. The entire system is moved by a wheel connected to a small stepping motor controlled by a timer, which regulates the interval and duration of movement. This movement tracks the sun, keeping the position of the focus fixed in the center of the ring. Some optical analysis can be performed based on Fig. 4. The real dimensions of all the mirrors and the angles they
Fig. 3. Detailed view of the focal ring, with holes for attaching measurement or experimental accessories.
1002
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006
The power in the different stages of reflections is the following: P r ¼ I r S1 ¼ RI 0 cos 45 S1
ð3Þ
P f ¼ I f S2 ¼ R2 I 0 cos 45 sin bS2
ð4Þ
where R is the mirror reflectivity, I0 is the direct solar irradiance, Ir is the irradiance reflected by the mirrors in the first set, If is the irradiance reflected by the mirrors in the second set, Pr is the power reflected by the mirrors in the first set, Pf is the power reflected by the mirrors in the second set, S1 is the mirror surface of the first set (S1mirror number of mirrors), S2 is the mirror surface of the second set (S2mirror number of mirrors). Table 3 gives Pr/R I0 and Pf/R2 I0 per ring. Adding up the contributions from each row, Pf can be expressed as:
Fig. 4. Geometric drawing for theoretical analysis.
P f ¼ 39445R2 I 0 ðI 0 in W=cm2 Þ form with the optical axis were measured. Table 1 gives the number of mirrors per ring in the first set, where ring 1 is the closest to the center, as well as the surface of each mirror and the effective collecting surface in each ring, taking the cosine factor into account as follows:
So, for a typical mirror reflectivity R = 0.8 with a direct solar irradiance of 0.1 W/cm2, the total power in the focus of the concentrator is 2.5 kW. The concentration factor can also be calculated. The irradiance in each stage can be expressed as:
S1effective; ring ¼ ðnumber of mirrors per ringÞ ðsurface of each mirrorÞ cos 45
I r ¼ RI 0 cos 45
ð6Þ
I f ¼ C mirrors2ndgroup R2 I 0 cos 45 sin b exp½2ðr=2rÞ2
ð7Þ
ð1Þ
From the data in Table 1, adding all the contributions from the third row, the effective collecting surface is 5.42 m2, so for a typical irradiance of 1000 W/m2, the collected power is 5.42 kW. However, as the rays are reflected twice and the incidence is not normal, mirror reflectivity losses and the different angles have to be considered. Table 2 lists the geometric data for the second set of mirrors (see Fig. 5). In this case, the equation for S2effective per ring is:
Table 4 lists Ir and If at the focal point (r = 0), assuming that the mirrors in the second set have a concentration factor of 8. The contribution of each ring to the concentration factor must be corrected for the angle formed by the optical axis and the corresponding ring (1–9), which has been labeled the h angle. Adding up the contributions from all the rings, the irradiance at the focus as a function of mirror reflectivity and the direct solar irradiance can be expressed as:
S2effective per ring ¼ ðnumber of mirrors per ringÞ ðsurface of each mirrorÞ sin b
ð5Þ
I focal ¼ 1910R2 I 0 ðwith I 0 in W=cm2 Þ
ð2Þ
ð8Þ
Table 1 Effective surface of the first set of mirrors. Ring
1
2
3
4
5
6
7
8
9
Number of mirrors S1mirror (cm2) S1effective per ring (cm2)
16 186 2104
24 186 3156
32 186 4209
40 186 5261
48 186 6313
56 172.5 6831
64 172.5 7806
72 172.5 8782
80 172.5 9759
Table 2 Effective surface of the second set of mirrors. Ring
1
2
3
4
5
6
7
8
9
Number of mirrors b angle S2mirror (cm2) S2effective per ring (cm2)
16 50 168 2059
24 51.875 168 3172
32 53.75 168 4335
40 55.625 156 5149
48 57.5 150 6072
56 59.375 150 7228
64 61.25 144 8080
72 63.125 144 9248
80 65 144 10,441
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006
1003
Table 4 Contribution by each row to irradiance in the different stages. Ring
1
2
3
4
5
6
7
8
9
Ir/(R I0) h If/(R2 I0) Ifocal/(R2 I0)
11 8 69 68
17 11 107 105
23 14 146 142
28 17 187 179
34 20 229 215
40 23 273 251
45 26 317 285
51 29 363 317
57 32 410 348
Fig. 5. Geometrical drawing of irradiance calculation in the focal plane.
For a reflectance of 0.8, the concentration factor is 1222 (i.e. the power density in the focus of the concentrator is 1222 times the direct solar irradiance), and for a reflectance of 0.7, the concentration factor is 936. 2.2. Measuring the system Gardon-type radiometers were used to map power density (Figs. 6a and 6b). These sensors are basically a differential thermocouple that measures the temperature difference between the center and the circumference of a thin circular foil disk (Gardon, 1953). The signal sent by the sensor is simply measured with a voltmeter (accuracy ±0.1 mv). The direct solar irradiance was measured continuously throughout testing with a Kipp and Zonen CH1 pyrheliometer (accuracy ±0.5%) and PT100 temperature sensor to correct the irradiance measurement. The Vatell Corp. is the only company currently manufacturing this kind of radiometer. They come with two types of black coating depending on the flux measurement range, ZynolyteÒ below 3500 kW/m2, and colloidal graphite above that. In this work, both types were used, taking into account the different spectral emissivity of the two coatings. However, the spectral distribution used in calibration is different from the solar spectrum, overestimating solar flux measured as a result. Therefore, when measuring concentrated solar radiation with such sensors a correction must be made. For radiometers with the ZynolyteÒ coating, the calibration constant correction factor is 0.965 and for colloidal graphite coating, it is 0.782 (Ballestrin et al., 2003). The accuracy quoted by Vatell for the radiometers was ±3%. The experimental device shown in Fig. 7 was designed to place the radiometers and move them around the focus. An elevator type scheme was chosen to provide movement on three axes. Movement along the optical axis (Z axis) was
Fig. 6a. Commercial Gardon-type radiometer.
Fig. 6b. Detail of the black radiometer coating.
electronically controlled, and movement on the XY plane was done manually. This device was attached to one of the holes in the focal plane delimiting ring in such a way that the positions of x and y in the [30, +30] interval could be measured with zero as the focal point. A polished stainless steel tube with holes in it to facilitate heat dissipation was connected to this device. The radiometer was
Table 3 Power contribution of each row of mirrors. Ring
1
2
3
4
5
6
7
8
9
Pr/(R I0) Pf/(R2 I0)
2104 1456
3156 2242
4209 3066
5261 3641
6313 4294
6831 5111
7806 5713
8782 6539
9758 7383
1004
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006
Fig. 9. Qualitative tests of two different specimens. (A) A 20 mm-dia. hole in a 0.5 mm-thick steel sheet after 40 s. (B) A 60 mm hole in a 1 mm-thick aluminum sheet after 10 s.
inserted in the tube, with cooling tubes shielded from direct radiation and insulated from heat with alumina (low thermal conductivity) and highly reflective aluminum foil (Fig. 8). 3. Results and discussion When all the mirrors had been properly focused, some qualitative tests were made. The first measured the temperature on a steel sheet used for focusing the mirrors with a Type K thermocouple. Temperatures up to 1500 K were recorded under good solar irradiance conditions. The second qualitative test was melting two different metals. A 0.5 mm-thick steel sheet was melted in 30–40 s, making a 20 mm-dia. hole in the center. A 1 mm-thick aluminum sheet was melted in just 10 s making a 60 mm-hole in the center. Both tests showed good beam symmetry, as seen in Fig. 9. As mentioned above, two types of radiometers were used for measuring solar power density, one with a colloidal graphite coating labeled 8170 and the other with a ZynolyteÒ coating labeled 8168. Fig. 10 shows the varia-
Fig. 8. Experimental setup used for the radiometer housing.
Power density (kW/m2)
Fig. 7. Radiometer housing device designed.
φpow=(1040
1250 1200 1150 1100 1050 1000 950 900 850 800 750 700 650 0.70
0.75
0.80
dir
0.85
0.90
0.95
1.00
2
Direct solar irradiance (kW/m ) Fig. 10. Direct irradiance vs. power density measured with two different radiometers. The concentration factor is 1040.
tion in solar power density in the focus (/pow is the power density and Idir is the direct solar irradiance). Experimental data from both radiometers with and without correction are shown. Only if the correction factor (Section 2.2) for the spectral emissivity of the two coatings is taken into account can the data from both radiometers be compared and a linear fit performed. These results show that the measured concentration factor is 1040 ± 60, which is quite near the one calculated with Eq. (8), giving an idea of the focusing quality, although it could still be slightly improved. Both types of radiometer were measured several times at nearly 300 different points all around the focal point, including the three axes, for a complete description of the power density distribution in the focus. When x had been set, a scan was made of the Z axis. When the scan was finished, x was reset and the Z axis was scanned again, and so on. The data for the Z = 0 focal plane can be fitted to a Gaussian curve (Fig. 11). Due to the good beam symmetry, this fit can be extrapolated to the whole space, and the power density in the focal plane can be expressed as:
/pow
r 2 ¼ C I dir exp 2 2r
ð9Þ
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006 Gaussian fit Experimental data
Normalized power density
1.0 0.8
C90
0.6 0.4 0.2 0.0 -80
-60
-40
-20
0
20
40
60
80
X axis (mm) Fig. 11. Normalized power density distribution in the focal plane and its Gaussian fit.
Relative power density
1.0 0.9 0.8 0.7 0.6 0.5 0.4 -100 -80
-60
-40
-20
0
20
40
60
80
Distance to the focal point along the optical axis (mm) Fig. 12. Power density variation along the optical axis. Positive axis faces the sun.
where /pow is the power density, C is the concentration factor of 1040, Idir is the direct solar irradiance, r is the distance to the focal point in mm and r is the standard deviation, which is (20 ± 2) mm. Defining the focal spot as the region where the power density is at least 90% of the maximum, it may be seen from the graphs that the diameter of the focal spot of this equipment is 20 mm. Variation of power density along the optical axis is shown in Fig. 12. It may be seen that there is nearly no variation in an interval of ±20 mm from the focal point. This
1005
feature makes the concentrator beam very suitable for application to materials treatments, including specimens larger than those previously treated with other concentrating technologies. The new concentrating system and the systems described in the introduction are compared in Table 5. As the main application of the new device is materials treatment, only the focus diameter and the power density need to be considered. The DRFL has a larger focus diameter than the Fresnel lens, although the power density is substantially lower. Nevertheless, the temperature achieved with the DRFL is high enough to conduct a wide variety of materials treatments. According to the table, the power density and focus diameter are considerably lower than other solar concentrators (central receivers, parabolic mirrors or solar furnaces), but the scale of the DRFL is a good advantage which should be taken into account. Another advantage over parabolic dishes is that due to the position of the focus, samples can be positioned almost face up. This new solar device is a good alternative for solar materials treatment mainly due to its small-scale and the focus diameter. The concentrator presented is only a prototype and can be improved in several respects. As mentioned above, each mirror is positioned manually, so how accurately this is done strongly affects the results, particularly the concentration factor. As mirrors get dirty, the concentration factor decreases, but when they are cleaned, the mirrors may be moved slightly out of position, losing focusing accuracy. A compromise must therefore be reached between cleaning and focusing. Two months between consecutive cleanings is good enough to keep focusing accurate and a good concentration factor. In future designs, the diameter of the focus could be reduced by changing the mobile mirror quality to increase the power density. A much better concentration factor would be achieved this way, although the cost of the equipment would increase. The optics could also be enlarged to increase system power, but one of the main advantages of the DRFL, which is its small-scale, would then be lost. 4. Conclusions A new low-cost and easy-to-use, small-scale solar concentrator has been completely described and characterized.
Table 5 Comparison of different solar concentrating systems. Kind of installation
Examples
Focus diameter (mm)
Power density (kW/m2)
Central tower and field of heliostats Parabolic mirror Solar furnace Fresnel lens Double Reflection Fresnel Lens
Almerı´a (PSA-CIEMAT), Themis (IMP-CNRS), Alburquerque, (Sandia National laboratories) Almerı´a (PSA-CIEMAT), Schengen, Suiza (PSI) Almerı´a (PSA-CIEMAT), Odeillo (IMP-CNRS). PSI Madrid (CENIM-CSIC), Ciudad Real (ETSII-UCLM), Lisboa (IST) Madrid (CENIM-CSIC)
1.5–3
700–3000
1–120 5–230 8 20
2000–4000 2500–17,000 2600 1000
1006
J. Llorente et al. / Solar Energy 85 (2011) 1000–1006
Flux mapping was done with Gardon-type radiometers and an elevating device to move the sensor around the focus. Two types of Gardon-type radiometers (ZynolyteÒ and colloidal graphite) were used showing good agreement with other studies on solar flux measurement. The concentration factor was 1040, with a 20 mm focal diameter at 90% power. The power density achievable and the focal diameter make this equipment suitable for materials treatments. Acknowledgements This work was funded by the Spanish Ministry of Education and Science Under Project ENE2004-07502, “High Temperature Concentrated Solar Energy Applications”. The new concentrating system described in this work was designed, patented and manufactured by Cruz and Bomant S. L. References Ballestrin, J., Ulmer, S., Morales, A., Barnes, A., Langley, L.W., Rodriguez, M., 2003. Systematic error in the measurement of very high solar irradiance. Solar Energy Materials & Solar Cells 80, 375– 381. Can˜adas, I., Martinez, D., Rodrı´guez, J., Fernandez-Gonzalez, B.J., Va´zquez Vaamonde, A.J., 2006. Procesamiento en lechos fluidizados calentados con energı´a solar concentrada. In: Proceedings of XII Congresso Iberico e VII Ibero Americano de Energia Solar Lisbon, Portugal. Costa Oliveira, F.A., Shohoji, N., Cruz Ferna´ndez, J., Guerra Rosa, L., 2005. Solar sintering of cordierite-based ceramics at low temperatures. Solar Energy 78, 351–361. Ferriere, A., P Rodriguez, G., Sobrino, J.A., 2004. Flux distribution delivered by a fresnel lens used for concentrating solar energy. Journal of Solar Energy Engineering – Transactions of the ASME 126 (1), 654– 660.
Ferriere, A., Sa´nchez Bautista, C., Rodrı´guez, G.P., Va´zquez, A.J., 2006. Corrosion resistance of stainless steel coatings elaborated by solar cladding process. Solar Energy 80 (109), 1338–1343. Flamant, G., Ferriere, A., Laplaze, D., Monty, C., 1999. Solar processing of materials: opportunities and new frontiers. Solar Energy 66 (2), 117–132. Gardon, R., 1953. An instrument for the direct measurement of intense thermal radiation. Review of Scientific Instruments 24 (5), 366–370. Guesdon, C., Alxneit, I., Tschudi, H.R., Wuillemin, D., Petrasch, J., Brunner, Y., Winkel, L., Sturzenegger, M., 2006. PSI´s 1 kW imaging furnace – A tool for high – temperature chemical reactivity studies. Solar Energy 80 (10), 1344–1348. ´ Gallagher, J., Winston, R., Sagie, D., Lewandowski, A., Bingham, C., O 1991. Performance characterization of the SERI high-flux solar furnace. Solar Energy Materials 24, 550–563. Neumann, A., Groer, U., 1996. Experimenting with concentrated sunlight using the DLR solar furnace. Solar Energy 58, 181–190. Riskiev, T.T., Suleimanov, S.K.H., 1991. Double mirror polyheliostat solar furnace of 1000 kW thermal power. Solar Energy Materials 24, 625–632. Rodrı´guez, G.P., de Damborenea, J.J., Va´zquez, A.J., 1997. Surface hardening of steel in a solar furnace. Surface and Coatings Technology 92, 165–170. Sa´nchez Olı´as, J., Garcı´a, I., Va´zquez, A.J., 1999. Synthesis of TiN with solar energy concentrated by a Fresnel lens. Materials Letters 38, 379– 385. Sierra, C., Va´zquez, A.J., 2005. High solar energy concentration with a Fresnel lens. Journal of Materials Science 40, 1339–1343. Sierra, C., Va´zquez, A.J., 2005. NiAl coatings on carbon steel by selfpropagating high-temperature synthesis assisted with concentrated solar energy: mass influence on adherence and porosity. Solar Energy Materials and Solar Cells 86, 33–42. Sierra, C., Va´zquez, A.J., 2006. Dry sliding wear behaviour of nickel aluminides coatings produced by self-propagating high-temperature synthesis. Intermetallics 14, 848–852. Va´zquez, A.J., Rodrı´guez, G.P., De Damborenea, J., 1991. Surface treatment of steels by solar energy. Solar Energy Materials 24, 751– 759.