## Heat loss coefficient

The present stagnation temperature studies allow determining the heat loss coefficient and the heat capacity of a collector. In this approach the collector is not necessarily connected to an operating heat removal system (e. g. water). This implies that the absorbed solar irradiance AcIG (ra)e is equal to the heat loss Ac UL (Tabs, m — Ta) plus the change in internal energy Ac Ce(dTabs, m/dt) of the collector:

Ac is the active collector area, IG the global solar irradiance, (ra)e is the effective transmittance — absorptance product, UL the heat loss coefficient, Tabs, m the mean absorber temperature, Ta the ambient temperature and Ce is the effective heat capacity of the collector per unit area. In the following (Tabs, m — Ta) will be referred to as AT. The solar energy absorbed by the collector per unit area can also be written in terms of direct and diffuse components,

IG (Ta)e = Ig ( Ta)en [kbKe (db) + kdKe (6d )] (2)

The term (ra)e in Eqn. 2 has been replaced by the effective transmittance-absorptance product at normal incidence (ra)en multiplied with the incidence angular modifier K6(6). kb and kd are the fractions of the global irradiance related to direct beam radiation and diffuse radiation respectively. 6, is the incidence angle of the direct beam radiation, while 6d is the equivalent angle of the diffuse radiation. The incidence angle of the direct beam can be calculated for any time of a day from formulas given in [15]. Duffie and Beckman [15] also provide formulas for the effective incidence angle of the diffuse radiation. Rearranging Eqn. 1 and inserting Eqn. 2 gives:

Ul = [M6(6b) + к„Кб(б„)]- Ce ^ (3)

In general, the U-value is assumed to have a linear dependence on AT as shown in Eqn. 4.

Ul = U1 + U2 AT (4)

In order to predict the fraction of the solar irradiance that is absorbed in the collector, previous knowledge about (ra)e is used. This makes it possible to identify both U1 and U2 by plotting the right side of Eqn. 3 against AT. The only unknown parameter on the right side is the heat capacity. When the heat capacity term is set to zero, the U-value is apparently larger before solar noon than after solar noon. The heat capacity is chosen so that the derived U-values before and after solar noon coincide. This approach to graphically determine the heat capacity will be shown in Section 5. Ui and U2 are then determined by using a least square fit (minimizing the sum of the squared residuals).

It is emphasized that there is a difference between the coefficients U1 and U2 in the present paper and c1 and c2 in EN 12975 [16]. The relation between the two constants is c1 = F’ U1 where F’ is the collector’s efficiency factor. F’ is close to unity for the present collectors in the present study (and generally for most polymeric collectors), hence the difference disappears.

## Preliminary Design Forces

4.1. Calculus Relations

Three coordinate systems can be defined (fig. 7), resulting by consequent rotations with angles у and p. According to the three coordinate systems, the components of weight and wind loads are:

Gz =-G; Gxl = Gz sin у; Gyl = 0; Gzl = Gz cos у;

Gx2 = Gxi = Gzsin y ; Gy2 = Gzisin P = Gzcos y sin P; Gz2 = Gzicos P = Gzcos y cos P; (1)

 i1

Fig. 8. Calculus diagram for daily axis forces.

Table 3. Loads on the A axis.

 Point Forces Moments x2 У^ Z2 x2 Ук. Z2 A Gx2 Gy2 + FC cos^-Pe ) Gz2 + Wz2 + FC sin^-Pe ) 0 M y 2 — Gx 2 ei 0 Fx2 Fy2 Fz2 A’ Gx2 Gy2 + FC cos^-P e ) Gz2 + W + Fc sin(ф-Pe) , / / / 2 + My 2 — G x2 Є1 + ll 2 A’’ 0 Gy2 + FC COs(ф-P e ) Gz2 + W + Fc sin^-Pe) / / / 2 My 2 — Gx 2 ei ll 2

Figure 9 presents the calculus diagram for the reactions in the daily rotational axis and in the linear actuator. The reaction in the linear actuator is calculated with relation

Table 4 presents the loads on the D axis and the forces on the bearings.

Table 3. Loads on the D axis.

 Point Forces Moments Xl Уі Zl MDx1 MDv1 MDz1 D Gx1 + FE cos(y-Ye ) Wyi Wz1 + Gz1 + FE sin(y — Ye ) — Wy1 (Єз + e1 cos P) — (Gz1 + Wz1)e1 sin p 0 Mz1 + Gx1e1 sin P Fxi Fyi Fzi D’ Gx1 + FE C°s(y-Y e ) + 2 +mda 12 Wyi Wz1 + Gz1 + FE sin(y-Ye ) + 2 + MDx1 І2 / / / D’’ Gx1 + Fe cos(y-Ye) 0 Wz1 + Gz1 + Fe sin(y-Ye) / / / 2 мол 2 2 MDx1 І2 The reactions on the bearings of t ie D axis are:

Figures 8, b and 9, b present the extreme positions of each linear actuator mechanism. The optimal design of a linkage mechanism with linear actuator is looking for symmetrical extreme positions C1 (E1) and C2 (E2) of the C (E) link, relative to the less loaded position with ф (у) = 90°, in order to obtain proper ф (у) pressure angles. The following relations have been established:

4.2. Results

Loads on the elements of the equatorial tracking system with linear actuator have been calculated for the six loading cases and for the six wind cases. The dimensions of the tracking system, involved in calculus are: l = 1480 mm; e1 = 40 mm; e2 = 85 mm; a = 160 mm; l1 = 600 mm; e3 = 80 mm; e4 = 140 mm; d = 125 mm; l2 = 50 mm.

Figure 8 presents the reactions on the bearings and on the linear actuators for two loading cases (a — loading case 1 and b-loading case 6 — see table 2), for which the biggest reactions result. There have been considered the possible wind cases, numbered from 1 to 6, presented in table 1.

Figure 11 presents the reactions on the bearings and on the linear actuators for all six loading cases (see table 2). There have been considered the wind cases 2 (a) and (4), presented in table 1, for which the biggest reactions result.

3. Conclusion

Based on the results presented above, the following conclusions can be drawn:

• Wind cases 2 (for front wind) and 5 (for back wind) give bigger radial loads for D axis and axial force in the seasonal actuator E; The effect of wind cases on the other loads is irrelevant (fig. 10), even if the maximum load on the A axis is given by wind case 4;

• The bigger axial load on the screw’s actuator C is maximum for loading cases 4 (tension) and 2 (compression) and wind case 2 (fig. 11, a); The bigger axial load on the screw’s actuator E is maximum (tension) for loading cases 1 and 2 and wind case 2 (fig. 11, a).

References

[1] R. N. Clark, B. D. Vick, Performance Comparison of Tracking and Non-Tracking Solar Photovoltaic Water Pumping Systems, Presentation at the 1997 ASAE Annual International Meeting Minneapolis, Minnesota (1997). http://www. cprl. ars. usda. gov/REMM Publishers. htm.

[2] D. Diaconescu, I. Visa, B. Burduhos, V. Dinicu, The Incidence Angles of the Trackers Used for the PV Panels’ Orientation. Part I: Equatorial Trackers, International Conference on Economic Engineering and Manufacturing Systems RECENT, Vol. X (2007).

[3] A. Roger, J, Messenger, Photovoltaic Systems Engineering (2004), CRC Press, Boca Raton.

[4] C. I. Co§oiu, A. Damian, R. M. Damian, M. Degeratu, Numerical and experimental investigation of wind induced pressures on a photovoltaic solar panel (2008) International Conference on Energy, Environment, Ecosystems and Sustainable Development, Algarve, Portugal.

[5] NP-082-04. Eurocode 1. Design Code. Bases of Design and Actions on Buildings. Action of Wind. Monitorul Oficial al Romaniei.

## Sensitivity analysis

Table 2. Material data and allowable stress for Copper and Aluminium.

Material Outs [MPA] ae [MPA] Op0.2 =Ox500,p0.2 [MPA] Ox500,aiiow [MPA]

Al 95 50 35 35

Cu 220 70 50 50

Allowable levels of stress in glass are counted with time dependency and distribution of brittleness and by these two reasons the stress in a glass when used as a construction material never shall exceed 8-9 MPa [3].

A safety factor defined as (4) will be used

Sf =—————— (4)

^T 500

Since a construction never will be stronger than its weakest point we also defined a weakest safety factor as in (5).

Sf weakest = min(Sfg, Sfp, Sft) (5)

2.3 Modelling

As a modeling tool we used a finite element analysis program called “MSC Marc Mentat” version 2005r2 (32bit). A solar collector has double symmetry axis, therefore only one quarter of the collector were modeled. The model was set up with thick shell Quad 4 elements, called element type 75 in the program. Thick shell elements using Mindlin’s shell theory is common used in applications like these

[4] . The elements in the mesh were up to 20*20 mm.

3. Results

A sensitivity analysis where done in order to see which factors had most affection of the stresses in the material. The sensitivity is a dimensionless number describing the response of the disturbance, e. g. if you make a disturbance of 1 % and the response is -1.9 the stress in that particular case will be reduced by 1.9 %. Material in absorber was copper. The result is compiled in to table 3.

The x and y both have trends that seem complex and needs more analysis for understanding.

z is more or less proportional i. e. all trends ~ 1.

Selection of material and thickness in absorber affect the price / thermal performance relation. That relation will not be questioned in this article, instead geometry properties of tubes and absorber will be kept constant on common used values described in table 1.[5]

tg have trends that seem complex and needs more analysis for understanding.

Eg will not be further investigated since there are no real good alternatives to soda glass with better price / thermal performance relation.

 Disturbed factor Unaffected values P P gas ^glassmax ^absmax ^tubemax x 1440 mm -0.46 0.26 0.05 -0.45 y 1400 mm -1.90 -0.36 -0.31 0.24 z 10 mm 1.03 0.91 0.89 1.03 Dtt 120 mm -0.68 -0.86 -0.96 0.54 0t 12 mm 0.36 0.41 0.36 -1.81 tg 4 mm 0.56 -0.75 -0.55 0.55 tp 0.25 mm 2.10 1.23 1.18 2.08 tt 0.85 mm 0.09 0.14 0.11 -0.82 Eg 69 GPa 0.16 -0.18 -0.15 0.16 Ep 118 GPa 0.72 1.02 1.06 0.71 Et 118 GPa 0.09 0.14 0.11 0.10
 Table 3. Sensitivity of different disturbances

 Trends

## Fluid flow investigations

As mentioned above, the properties of the small test absorber are not directly scalable to larger absorbers. In [6] Treikauskas compares the simulated pressure drop of an optimized roll-bond absorber (harp absorber, 985 mm x 1925 mm) with state-of-the-art absorbers and with the measurement results of the small FracTherm® test absorber (590 mm x 1000 mm). However, it is difficult to compare the pressure drop of absorbers with different sizes. The pressure drop of a large FracTherm® absorber cannot yet be anticipated, and for a given volume flow it will probably be different from the small one. Therefore it will be very interesting to compare the large FracTherm® absorber to be developed with another large, optimized roll-bond absorber with a different channel design.

The development of the header channel is an important task with respect to connectivity and pressure drop. In order to have a first estimation of the pressure drop in a flat, wide header channel, simple analytical calculations of 1 m long channels with different rectangular cross sections were carried out. The channel width b was varied from 50 mm to 150 mm, the height h from 2.95 mm (height used in small test absorber) to 4 mm. The results for different volume flows Q are shown in Fig. 5. The change from laminar to turbulent flow is obvious: the pressure drop rises with Q according to a linear function in the laminar region and approximately according to a power function in the turbulent region. Fig. 6 shows both the pressure drop for different rectangular cross sections for a volume flow of Q=0.36 m3/h and the cross sectional areas. It can be seen that same cross sectional areas (which means same fluid volume and thus thermal capacity) lead to different pressure drops. The example in the diagram reveals a pressure drop difference of about 24 % for A=300 mm2 (see dashed arrows). It is evident that for a given cross sectional area the pressure drop becomes lower with increasing h/b ratio. But it is a question of technical feasibility whether the height of the channels can really be increased.

2. Conclusion

The results of the FracTherm® test absorber investigations carried out earlier are not directly scalable to standard size collectors. Therefore the main challenges of the European project BIONICOL are the further development of the FracTherm® program, the development of appropriate heat transfer fluids in order to prevent corrosion, the adaptation of a glass batch coating plant for coating solar absorbers, the development of concepts for header channels and the interconnection of collectors and finally the field tests to be carried out for one year in different sites in Europe. Some first rough calculations of the pressure drop in rectangular channels were already carried out. They can serve as a basis of dimensioning roll-bond header channels with a given maximum channel height.

3. Acknowledgement

The contract for the BIONICOL project was not yet signed when this paper was written. This is the reason why the project partners are not mentioned by name. The consortium applied for the project in the call FP7-ENERGY-2007-2-TREN within the Seventh Research Framework Programme (FP7) of the European Commission.

References

[1] V. Weitbrecht, D. Lehmann, A. Richter, Flow distribution in solar collectors with laminar flow conditions. Solar Energy 73(6), pp. 433-441, 2002

[2] M. Hermann, FracTherm — Fractal hydraulic structures for energy efficient solar absorbers and other heat exchangers. Proceedings, EuroSun 2004, Freiburg, Germany, 20-23 June 2004, Volume 1, pp. 332-338

[3] M. Hermann, Entwicklung des FracTherm-Absorbers — Simulationen und Experimente. Proceedings, 15. Symposium Thermische Solarenergie OTTI, Bad Staffelstein, Germany, 27-29 April 2005, pp. 94-99

[4] M. Hermann, (2005). Bionische Ansatze zur Entwicklung energieeffizienter Fluidsysteme fur den Warmetransport. Dissertation, Faculty of Mechanical Engineering, Universitat Karlsruhe (TH)

[5] C. Mattheck, Teacher tree: The evolution of notch shape optimization from complex to simple. Engineering Fracture Mechanics 73 (2006), pp. 1732-1742

[6] F.-D. Treikauskas, W. Zorner, V. Hanby, Volumetrische Absorber: Die neue Generation von Solarabsorbern in Theorie und Praxis. Proceedings, 18. Symposium Thermische Solarenergie OTTI, Bad Staffelstein, Germany, 23-25 April 2008, pp. 182-187

## Losses of Solar Thermal Collectors in General

In steady state conditions the useful heat of a solar thermal collector Quse, coll in W can be simply expressed by

Optical losses [W]

Solar radiation energy is incident on the collector at its aperture area Aaperture. This radiation Gglob, i consists of direct and diffuse radiation and is measured in the plane of the collector aperture. Two types of losses occur: optical losses and thermal losses. The terms used in eq. (1) are explained in detail in [5] and [6]. It depends on the construction of the collector and on its working temperature which kinds of losses are dominant and therefore have to be reduced.

In general, optical losses occur at the transparent cover of the absorber due to absorption and reflection as well as at the absorber by reflection. If reflectors are used, additional optical losses due to absorption and diffuse scattering of the reflected radiation have to be taken into account.

The optical losses include all incident solar radiation that does not reach the absorber of the collector. After absorption of the remaining part of radiation, convective losses occur due to natural convection in the gap between the absorber and its cover as well as due to forced convection by wind passing the cover. Conductive losses appear in the gap between the transparent cover and the absorber as well as through the back insulation or the frame of the collector. Both types of losses can be nearly eliminated by placing the absorber inside of a vacuum. Radiative losses of the absorber rapidly grow with increasing working temperature of a collector.

Radiative losses can be reduced by selective coating. Further reduction requires a relative small absorber area compared to the area of the aperture. Thus concentrators have to be applied to bundle the incident light on the absorber. In the SHIP Task different approaches

to reduce losses were considered and new types of collectors were developed.

## Conclusions and Further Works

The „Ray-Tracer” software uses the Ray-Tracing method to study some physical phenomena and is useful to study and simulate the mechanisms that take place in the solar radiation concentration process from the non-imaging photovoltaic concentrators. The paper presents results of simulated experiments regarding the concentration of the radiation by a paraboloidal photovoltaic concentrator installed on the roof of a house in the insolation conditions of a clear sky day.

 Fig.3 Fig.4
 Fig. 7 Fig.8

The geometric characteristics of the paraboloidal photovoltaic concentrator are: parameter p = 200 mm, input aperture radius R = 50 mm, height h = 62.5 mm, focal distance f = H0 = 10 mm, geometric concentration factor Cgeom = 6.25. The photovoltaic cell is placed in the focal plane of the paraboloid and has the radius r = 20 mm. The photovoltaic concentrators are placed on the roof oriented South, у = 0, and the inclination is equal with the latitude of the place 5 = 45 deg. The maximum value of the incidence angle on the input aperture to which the concentration produces is 6max = 30.06 deg. For the given case, the minimum value of the incidence angle is reached at noon in September, 6min = 1.507 deg. The theoretical factor of optical concentration is Coptic, teoretic = 3.98. For the given situation, the maximum optical concentration factor is Cmax = 3.57 in September at noon. The optical concentration factor, in the described experiments, is higher than 1 between 1030 o’clock and 14 o’clock. The maximum density of the solar radiant flow on the photovoltaic cell is reached in March, at noon, Brec = 2837 W/m2. The efficiency of the concentration varies between 12.31%(June) and 26.10% (September). A photovoltaic installation with the collecting area of 12.50 m2, with 5000 paraboloidal concentrators, provides monthly the electric energy quantity variyng from 71 kWh, in December, to 383 kWh, in March. The electric energy production satisfies the the need of a family and allows the monthly delivery, in spring, summer, autumn and winter of approximately 125 kWh to the national energetic system.

The following papers will refer to installations with paraboloidal concentrators at which the distance between the parabola’s peak and the photovoltaic cell is variable. The purpose of these papers is to determine the optimum position of the cell depending on the paraboloid’s peak so that the optical efficiency to be higher that 1 for a longer period of time.

Acknowledgments. This work was supported by the grant CEEX-247 References

[1] Swanson, R. M., Photovoltaic concentrators in Photovoltaic science and engineering edited by Luque, A. and Hegedus, S., Wiley, pp. 449 — 505, 2002.

[2] Bowden, S. B.E., A High efficiency photovoltaic roof tile, a thesis of University of New South Wales, April 1996.

[3] Brogren, M., Optical efficiency of low-concentratingsolar. Energy systems with parabolic reflectors, Uppsalo University, Sweden, 2004.

[4] Winston, R., Principles of Solar Concentrators, Solar Energy 16, pp. 89-95, 1974.

[5] McIntosh, K. R., Swanson, R. M., Cotter, J. E., A simple ray tracer to compute the optical concentration of photovoltaic modules, Progres In Photovoltaics: Research And Applications, 14, pp 167 — 177, 2006.

[6] Gray, J. L. , The physics of the solar cell concentrators in Photovoltaic science and engineering edited by Luque, A. and Hegedus, S., pp. 61 — 113, 2002.

[7] Glassner, A. , ed. An introduction to ray tracing, Academic Press, San Francisco, pp. 263—294, 2002.

[8] Emery, K., Measurement and characterization of solar cells and modules concentrators in Photovoltaic science and engineering edited by A Luque, A. and Hegedus, S., Wiley, pp. 701-753, 2002.

[9] Nilsson, J., Optical design and caracterization of solar concentrators for photovoltaics, Lund University, Licentiate Thesis, 2005.

[10] Danescu, Al., Bucurenciu, S., Petrescu, St., The utilisation of solar energy (in Romanian), Ed. Tehnica, 1982.

[11] Fara, L., Tulcan-Paulescu, E., Paulescu, M., Photovoltaic systems (in Romanian), Matrix, 2005.

[12] Paulescu, M., Schlett, Z., Practical aspects in the photovltaic conversion of the solar energy (in Romanian), Mirton, 2002.

[13] De Sabata, C., Luminosu, I., De Sabata, A., Palea, A., On the Design of a Solar, Partially Energetically Independent House in the Region of Banat, Bul. St. Univ. "Politehnica" din Timisoara, Transactions on Mechanics, 52(66),4, 2007, pp. 82-87, ISSN 1224-6077, .

[14] Luminosu, I., De Sabata, C., De Sabata, A., Theoretical and experimental researches over the posibility of realizing a solar house that is partially independent thermoenergetically (in Romanian), Buletinul AGIR, fondat 1918, ISSN 1224 — 7928, XII, nr. 3, iulie — septembrie, 2007, pp.31 — 44.

[15] Paulescu, M. , Algorithms for estimating the value of the solar energy (in Romanian), Matrixrom, Bucuresti, 2005.

[16] Luminosu, I., Zaharie, I., Costache, M., Damian, I., Optical concentrators in photovoltaics installations, Conferinta Nationala ”Instalatiile pentru Constructii §i Confortul Ambiental”,183 — 190, Timisoara, martie 2007.

[17] Luminosu, I., Zaharie, I., Costache, M., Damian, I., Ray-tracing — an analysis method of optical concentrators, Conferinta Nationala ’Instalatiile pentru Constructii §i Confortul Ambiental”, pp191 — 199 , Timisoara, martie 2007.

[18] Luminosu I., Nagy M., Mirrors and thermo-solar devices with radiation concentrators

produced at the Polytechnics University in Timisoara, Bulletins for Applied and Computer Mathematics, (PAMM), Budapest University of Technology and Economics, BAM Nr. 2173, pp. 135 — 146, 2004.

[19] **** Tehnical Report of CEEX-247, dec. 2007

## Effect of thermotropic layers on collector efficiency

In Fig. 1 the efficiency factor is plotted as a function of the absorber temperature for a collector without and with thermotropic overheating protection at a solar irradiation of 1200 W/m2 and ambient air temperatures of 0 and 30°C. The thermotropic layer exhibits a solar transmittance of 0.90 in clear state and of 0.10 in opaque state. The collector without a thermotropic layer reaches maximum

absorber temperatures of ~160°C and ~175°C at ambient temperatures of 0 and 30°C, respectively. It is observable that the stagnation temperatures can be controlled by the use of thermotropic layers. To achieve maximum absorber temperatures of 90°C, switching temperatures of the thermotropic film between 55 and 60°C are required. A slight effect of the ambient air temperature on the efficient working temperature range is discernible. At elevated ambient air temperatures of 30°C the efficiency drop is shifted to lower temperatures. However, the efficient working temperature exceeds 60°C even at an ambient temperature of 30°C. Thus the overheating protected collector is appropriate for domestic hot water and space heating applications.

 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 absorber temperature [°C]

Fig. 1. Collector efficiency versus absorber temperature of a solar collector with twin-wall sheet glazing and
black absorber (a=0.95, є=0.90) at a solar irradiation of 1200W/m2 and ambient air temperatures of 0°C and
30°C; solid lines: collector without overheating protection (no TTL); dashed lines: collectors with thermotropic

glazing (switching temperature: 55-60°C).

In Fig. 2 the effect of the thermotropic layers switching range on maximum absorber temperatures at a solar irradiation of 1000W/m2 and an ambient air temperature of 20°C is shown. The residual solar transmittance of the thermotropic layer in opaque state is varied between 0.20 and 0.60. The transmission of the layer in clear state remains constant at 0.85. Compared to layers exhibiting a solar transmittance of 0.90 in the clear state (Fig. 1), the collector efficiency is shifted to slightly lower values. As to layer design, this indicates that the solar transmittance should exceed 0.85 in the clear state. The collector without thermotropic overheating protection reaches absorber temperatures of about 160°C. It is observable that by applying thermotropic layers exhibiting a residual transmittance of 0.30 to 0.35 the absorber temperatures do not exceed 90°C. Layers exhibiting transmittance values above 0.35 lead to a successive increase of maximum absorber temperatures up to 130°C for a solar transmittance of 0.60 in the scattering state.

 20 30 40 50 60 70 80 90 100 110 120 130 140 150 absorber temperature [°C]

Fig. 2. Collector efficiency versus absorber temperature of solar collectors with twin-wall sheet glazing and black absorber (a=0.95, є=0.90), at a solar irradiation of 1000W/m2 and an ambient air temperature of 20°C; variation of switching performance of the thermotropic layers (TTL) (solar transmittance: 0.85 in clear state and 0.20 to

0.60 in opaque state).

## Coatings with paint

Figure shows a summary of the work done to produce selective paint. It is possible to observe high solar absorption values, but undesirably also high thermal emissivity. The paints obtained until the moment are not selective.

In the initial work with paints, the objective was to get good optical properties for paint with the organic pigment C6o/ C70. High solar absorption (95% and 96%) was reached. The problem was the emissivity, which is strongly dependent on coating thickness. With the coil method adopted for coating, the lower thickness achieve was 7pm, with 80% of emissivity and 95% of solar absorption. To reduce the thickness and consequently the emissivity, spray technique was tested and it was possible to achieve 4pm of thickness and emissivity of 74%, with 96% absorption.

 Fig. 6. Absorption variation with wavelength for different paint samples.

Without the possibility to reduce thickness to lower values with methods of easy application, it was also tested the incorporation of metallic pigments in the paint with 16% CVP of C60/ C70 pigment, considering that the thermal conductivity of metallic pigments would lower the emissivity values. Both copper pigment with average grain size between 63 and 90pm and stainless steel with average grain size of 3 pm were tested. The mix was done adding 16% of metallic pigment weight to the already prepared paint with C60/ C70 pigment. Figure 6 shows that this did not improve the paint behaviour in relation to emissivity.

Adding higher quantity of metallic pigment, about 50%, to the base of paint, without use of organic pigments, hoping to increment thermal conductivity of the coating and obtain lower emissivity,

independently of thickness coating, also did not improve the emissivity and, without organic pigment, the absorption decreased to 36%. The fact that the metallic pigment used, was stored for a long time (surface highly oxidized) could cause the observed behaviour. Also the surface shape of used pigments could explain the observed behaviour, since the surface contact area between metallic particles and the metallic substrate was not adequate to increment conductivity. These aspects will be explored in near future.

Topography of paint with organic pigment obtained by SEM (Fig.7.a) allows us to identify a granular morphology, with grains agglutinated by resin. It is visible agglomerates of small grains; which rough surface that can improve absorption.

Fig. 7. a) SEM (30000x) surface micrograph of paint with organic pigment. b) Surface
photography by optic microscope (45x) of paint with organic and Cu grains.

4- Conclusions

Optical properties of titanium oxide are strongly dependents of deposition parameters, and some of these are interrelated, which become very difficult to relate optical properties with change of each parameter, but it is possible to conclude that best values of absorber selectivity were obtained in dc mode and in pulsed dc mode with 200kHz, with oxygen flow rate changing between 0 and 2.5ml/min with adequate slope. Adequate slope depends of deposition rate which depends of deposition power, total pressure, oxygen partial pressure and pulsed frequency and all of these parameters are important, once that for solar absorber selectivity the final thickness and oxygen gradient concentration along of the film thickness are determinants. Best optical properties for oxide titanium sputtered films were 88% for solar absorption, with 7% of emissivity for deposition parameters of: pulsed frequency 200kHz, reverse time of 0.4ps, discharge current of 0.7A, argon flow rate of 50ml/min and oxygen flow rate changing from 0 to 2.5ml/min. The morphology of oxide titanium films is columnar, with columns oriented in direction of growing film, which seem to be continuous from the substrate to the top of the film. Subsequent immersion in solution with antocyanin didn’t show to improve solar absorption.

For paints, the results obtained until the moment weren’t satisfactory. The best couple values for solar absorption and emissivity were respectively 94%, and 74%. Emissivity is dependent on thickness of coatings and with the used application techniques, the minimum thickness reached was 4pm, not low enough to obtain infrared transparency. The effort to reduce emissivity of paints adding metallic particles were unfruitful, at least using for the shapes and sizes of metallic particles used. Surface topography shows grains agglutinated with binder.

Aknowledgements -To Fundagao para a Ciencia e Technologia by the financial support through the

referred research project POCTI/ENR/62660/2004 “Development of new spectrally selective coatings with

organic pigments for absorbers of solar collectors.”

References

[1] Project POCTI/ENR/62660/2004 “Development of new spectrally selective coatings with organic pigments for absorbers of solar collectors”, Fundagao para a Ciencia e Tecnologia.

[2] M. J. Brites, C. Nunes, S. Vieira, D. Quintino, Lopes Prates, J. Alexandre, V. Teixeira, M. J. Carvalho, Proceedings of CIES 2006-XIII Iberic Congress, and VIII Iber American Congress of Solar Energy, Lisboa, Portugal, 9-10 November 2006.

[3] A. J. Martins, C. Nunes, M. J. Brites, M. Lopes Prates, V. Teixeira, M. J. Carvalho, Journal of Nanoscience an Nanotechnology, Vol. 8, 1-5, 2008.

[4] C. Nunes, V. T eixeira, M. L.Prates, N. P.Barradas, and A. D. Sequeira, Thin Solid Films 442, 173 (2003).

[5] M. J.Brites, C. Santos, B. Gigante, S. Nascimento, H. Luftmann, A. Fedoro v, and M. Berberan-Santos, New J. Chem. 30, 969 (2006).

[6] S. Nascimento, M. J.Brites, C. Santos, B. Gigante, A. Fedoro v, and M. Berberan-Santos, J. Fluoresc. 16,

245 (2006).

[7] K. Hara, Y. T achibana, Y. Ohga, A. Shinpo, S. Suga, K. Sayama, H. Sugihara, and H. Arakawa, Sol. Energy Mater. and Sol. Cells 77, 89 (2003).

[8] M. K.Gunde, Z. C.Orel, and M. G.Hutchins, Sol. Energy Mater Sol. Cells 80, 239 (2003).

[9] Z. C.Orel, Sol. Energy Mater. Sol. Cells 68, 131 (2000).

[10] Z. C.Orel, Sol. Energy Mater. and Sol. Cells 68, 337 (2001).

[11] K. Zakrzewska, M. Radecka, A. Kruk, W. Osuch, Solid State Ionics,8404 (2002).

[12] S. G.Springer, PE. Schmid, R. Sanjines, F. Levy, Surface and Coating Technology 151-152 (2002) 51-54.

[13] J. Y. Kim, EBarnat, E. J. Rymaszewski, T. M. Lu, J. Vac. Sci. Technology A 19(2), 429-434, Mar/Apr 2001.

[14] V. Vancoppenolle, P. Y.Juan, M. Wautelet, J. P.Douchot, M. Hecq, Surface and coating technology 116­119 (1999) 933-937.

[15] S. G.Springer, PE. Schmid, R. Sanjines, F. Levy, Surface and Coating Technology 151-152 (2002) 51-54.

[16] P. S. Henderson, P. J.Kelly, R. D.Arnell, H. Backer, J. W. Bradley, Surface and Coating Technology 174­175 (2003) 779-783.

[17] R. D. Arnell, P. J. Kelly, J. W. Bradley, Surface and Coating Technology 188-189 (2004) 158-163.

[18] N. J. Cherepy, G. P. Smestad, M. Gratzel, J. Z. Zhang, J. Phys. Chem B 1997, 101, 9342-9351.

[19] H. Backer, P. S. Henderson, J. W.Bradley, P. J.Kelly, Surface and Coating Technology 174-175 (2003) 909-913.

[20] A. Belking, Z. Zhao, D. Carter, L. Mahoney, G. McDonough, G. Roche, R. Scholl, H. Walde, Society of Vacuum Coaters, 43rd Annual Techn. Conf. Proc.-Denver, April 15-20, 2000.

[21] Jindrich Musil, Jan Lestina, Jaroslav Vlcek, Tomas Tolg, J. Vac. Sci. Technology A 19(2), 420-424, Mar/Apr 2001.

[22] I. Safi, Surface and Coatings Technology, 127 (2000) 203-219.

[23] P. J. Kelly, C. F. Beevers, P. S. Henderson, R. D. Arnell, J. W. Bradley, H. Backer, Surface and Coating Technology, 174-175, 795-800, 2003.

[24] Sang-Wong Park, Jae-Eun Heo, Separation and Purification Technology 58 (2007) 200-205.

## Technical improvement of a small modular parabolic trough collector

Klemens Schwarzer, Jan Kroker, Markus Rusack

Fachhochschule Aachen / Campus Julich
Solar-Institut Julich (SIJ)

Heinrich-Mufimann-Str. 5, D-52428 Julich, Germany
Corresponding Author: schwarzer@sij. fh-aachen. de

Abstract

Throughout an earlier research project the Solar-Institut Julich (SIJ) developed a small sin­gle axis tracking modular parabolic trough collector with an evacuated absorber tube, a high — reflective aluminium mirror, an anti-reflective solar glass cover and a step motor drive with worm gear and tracking system. In consequence of the small aperture area of 2 m2 the collec­tor is lightweight and can be used for roof mounting. The aspired operating temperature level is 120 to 200 °C. Measurements and regression analysis have shown an overall effi­ciency of approximately 63 % at a working temperature of 160 °C (relating to 800 W/m2 of solar direct radiation).

In March 2007 the SIJ has started a further research project in association with four partners of the German industry in apparatus engineering, absorber technology and drive engineering. The main aim of the project is the improvement of the collectors’ technical characteristics and its thermodynamic performance. In addition the production costs are to be reduced. Fur­ther the project aims at a series production of the collector.

The project is state-aided by the German Federal Ministry of Education and Research. Keywords: small parabolic trough collector, collector deep drawing, collector improvement

1. Background

Throughout an earlier research project from 2003 to 2005 the Solar-Institut Julich (SIJ) developed a small single axis tracking modular parabolic trough collector, named PTC 1000. Possible appli­cations are the supply of process heat for hotels and hospitals, for industry applications and for the supply of cooling energy. With the end of the project the construction of the collector was finished and three prototypes were built.

Since March 2007 the SIJ tends to optimize the PTC 1000 prototype in a continuative project since weak points not only of constructional background had been detected throughout the testing phase. The main aim is to achieve a series-production readiness. The project is accomplished in collabora­tion with four partners of the German industry in apparatus engineering, absorber technology and drive engineering, which are Wallstein Ingenieur-Gesellschaft mbH, NARVA Lichtquellen GmbH + Co. KG, Ingenieurburo Annas & Partner GmbH and SMF Spanlose Metall Formung GmbH & Co. KG.

The project is divided into four work packages: In the first stage, which ended end of Octo­ber 2007, the weak points of the prototypes such as thermodynamic and technical deficiencies have been identified and a specification sheet for the new collector design has been elaborated. In the second stage concepts for the individual components have been worked out in detail and the new design concept of the collector has been decided at the end of June 2008. Currently, a detailed planning takes place, so that the production can start in September 2009. In the last work package the newly developed collector prototypes will be set up at the test facility of the SIJ and perform­ance tests will be carried out to evaluate the new collector design.

To show the demand of optimization regarding the collectors’ construction and thermodynamics, the features and the determined deficiencies of the prototype are described in the following.

## Experimental set-up

Solar collector efficiency has been measured with the indoor test rig shown in Fig. 1. It consists of a 12kW mercury lamp array, an integral thermostat plant, fluid temperature and air temperature probes, a pyranometer, an anemometer, a precision flow meter and a wind generator.

 Fig. 1. Test rig scheme.

This installation meets the requirements established in the European Standard EN 12975 to measure the efficiency of solar collectors.

Two prototypes of solar collectors have been manufactured using tabulators. In the first one, we have used a continuous twisted copper tape. The tape was placed all along the riser tubes of the harp. It had

0. 2 mm thickness and 5 mm width. The length of the 180° twist was 20 mm approximately. The second one was made using a steel chain as an insert. The chain had 5 mm width. It was placed in the same way as the first one, in the riser tubes. Both prototypes were constructed using the basis of a commercial collector of Isofoton, risers having an inner diameter of 7 mm. This way, the only difference between the prototypes and the standard design was the addition of tabulators.

The experimental sequence was as follows: i) efficiency test of the commercial collector, ii) efficiency tests of prototype 1 and iii) efficiency tests of prototype 2. In order to analyze the influence of water flow, we measured the efficiency of the prototypes at three different flows: 160 kg/h, 320 kg/h and 500 kg/h.

Each test was made during a whole day, and all tests were carried out in consecutive days. All of them were completed according to EN 12975.

2. Results

Table 1 shows a comparison between the efficiency of the standard collector and the turbulators prototypes (mass flow of 160 kg/h).

Table 1. Efficiency coefficients

 Standard Prototype 1 (twisted tape) Prototype 2 (chain) П0 0.753 0.782 0.773 a1 3.184 3.052 3.363 a2 0.0138 0.0176 0.0073

The main result is the 3% increase of q0 in the first prototype. However, there is also an opposite growth of the loss coefficients, ai y a2. In order to make the analysis, all three efficiency curves have been plotted in Fig. 2.

 Fig. 2. Efficiency curves.

The 3% gain in the left side of the curve seems to be reduced in the right side to 2%. Although prototype 1 is better, the second prototype equals its efficiency at non dimensional temperature T*=

0.07. We can confirm that there is a consistent efficiency increase along the curve when using tabulators.

Furthermore, we have made three different tests for both prototypes, at three different mass flows: 160 kg/h, 320 kg/h and 500 kg/h. The results of these tests are shown in Tables 2 and 3.

Table 2. Prototype 1. Efficiency coefficients at different mass flows

 160 kg/h 320 kg/h 500 kg/h П0 0.782 0.785 0.782 a1 3.052 3.45 3.16 a2 0.0176 0.012 0.013

Table 3. Prototype 2. Efficiency coefficients at different mass flows

 160 kg/h 320 kg/h 500 kg/h П0 0.773 0.779 0.78 a1 3.363 3.085 3.35 a2 0.0073 0.0219 0.014

It is observed that there is no significant variation in the efficiency in terms of mass flow, in any case. For both prototypes, the efficiency remains at approximately the same value.

The uncertainty of the efficiency curves has been estimated according to EN 12975 [4], and its value is ± 1.9%.

3. Conclusions

Experimental tests have demonstrated the suitability of using tabulators to improve solar collectors’ efficiency. A 2-3% efficiency increase can be obtained. Moreover, the insertion of twisted tapes has been reported to be a better option than the use of a chain.

There is no significant variation of the efficiency depending on mass flow when the two types of turbulators described are used.

A more detailed study to optimize the design of the tabulators will be done. However, the simplicity of the materials used and the efficiency enhancement obtained in this work, demonstrate that this solution is an adequate and suitable way of improving solar collectors.

References

[1] Duffie, Beckman. Solar engineering of thermal processes. Wiley-Interscience, 1980.

[2] P. Promvonge, S. Eiamsa-ard. Heat transfer behaviors in a tube with combined conical-ring and twisted-tape insert. ScienceDirect, Elsevier, 2007.

[3] S. Ray, A. W. Date. Friction and heat transfer characteristics of flow through square duct with twisted tape insert. ScienceDirect, Elsevier, 2002.

[4] UNE EN 12975. Sistemas solares termicos y componentes. Captadores solares. AENOR, 2006.