Energy Conversion and Management 52 (2011) 32413248

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Design and optical performance of a nonimaging Fresnel transmissive concentrator for building integration applications
Daniel Chemisana , Joan Ignasi Rosell
Applied Physics Section of the Polytechnic School (EPS), University of Lleida, 25001 Lleida, Spain

a r t i c l e

i n f o

a b s t r a c t
A transmissive Fresnel reector is designed to match the needs of building integration for concentrating photovoltaic (PV), thermal (T) or hybrid photovoltaic/thermal (PVT) generation. The device concentrates radiation toward a static receiver by means of an array of reectors which rotate collectively. All rotation axes are coplanar and parallel. A deep analytical ray tracing study has been made of the design characteristics and concentrator performance, thus determining the conguration which optimises efciency. Numerous ray tracing numerical simulations have been performed which contrast and support the analytical results. 2011 Elsevier Ltd. All rights reserved.

Article history: Received 1 December 2009 Received in revised form 30 July 2010 Accepted 11 May 2011 Available online 28 June 2011 Keywords: Building integration Fresnel Solar concentration Nonimaging/anidolic

1. Introduction At present, use of solar concentrator systems is mainly limited to large installations with devices of considerable size. Most of the manufacturers of solar concentrators are focusing their developments to these huge solar plants and there are very few examples in the market of middle or small scale installations. The medium or small scale installations would allow for a feasible integration of the concentrating solar systems into the buildings, leading to more distributed energy production scheme. Building Integrated Concentrating Photovoltaics systems may be installed either on the building faade or on the roof (which may be at or sloped) producing in each case a different visual impact. Depending on the type of device, the system may be integrated in such a way that it is unseen, plays some role in the architectural aesthetic or that it constitutes in itself an architectural concept [1]. The integrability of a concentrator, being it reective or refractive, depends on its concentration factor, C (dened as the ratio of the entrance aperture width to the exit aperture width). Concentrating systems with C > 2.5X use a system to track the sun, whereas systems with C < 2.5X can be static. However, in the long term static concentrators with higher ratios which make use of luminescence and photonic crystals may appear. Low concentrating ratio systems (C < 10X) are of particular interest for as they are of linear geometry and thus one tracking axis is sufcient for efcient operation [2].
Corresponding author.
E-mail address: daniel.chemisana@macs.udl.cat (D. Chemisana). 0196-8904/$ - see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.enconman.2011.05.006

Furthermore, concentrating photovoltaics is a feasible method to reduce the high initial cost of photovoltaic solar energy. Concentrating the solar radiation onto solar cells implies that the area of semiconductor devices is diminished by replacing part of it with a cheaper element (the concentrator). Considering that a higher concentration factor results in higher cost reduction, it can be seen that within the concentration range where single axis tracking may be used, the most desirable concentration factor is that which approaches the upper limit. For the aforementioned range of concentrations, good versatility is offered by systems which work using Fresnel reection. The technology which groups the great majority of the Fresnel solar concentrators is the one in which the tracking is achieved by the movement of the individual reectors whilst the receptor remains static opposite them. These systems are designed for power generation on an industrial scale, and in consequence this leads to a difcult direct building integration. By considering the elements sizing to be architecturally incorporated, this option may be easily installed on either at or inclined roofs. Installation on faades however presents certain problems: the reectors prevent light from passing into the building and the receiver must protrude outward from the building creating strain on the building structure and an anaesthetic appearance. The most important design within this group is the Compact Linear Fresnel Reector (CLFR). The CLFR system is used for the direct steam generation. Similar systems to the CLFR have been developed, being an example the solar collector Solarmundo [3]. The University of Lleida has constructed another such system with a PVT receiver in 2009 in collaboration with NUFRI corporation and Trigen Solar S.L.

3242

D. Chemisana, J.I. Rosell / Energy Conversion and Management 52 (2011) 32413248

Another type of Fresnel reective concentrators make the tracking functions by movement of the entire system using 2-axis actuators, such as the BiFres system developed at the University of Lleida (equipped with a photovoltaic-thermal PVT receiver), whose integration in buildings would be restricted to at (horizontal) roofs [4]. A new concept in the eld of Fresnel reection systems is the so called Non-Imaging Reective Lens (NIRL) concentrator, of which there are two types: the axially symmetric Ring Array Concentrator (RAC) and the linearly symmetric Slat Array Concentrator (SAC). These operate by using reectors to direct and concentrate light onto a receiver behind the optical element thus emulating a lens. The high concentration, RAC, requires two axis tracking, whereas the medium concentration SAC can be employed with either one or two axis tracking. This type of concentrator combines the high optical efciency achievable by mirrors with the exibility of design which is characteristic of lenses [5]. The principle drawback of these systems is that solar tracking is achieved by movement of the whole system, incurring the aforementioned restrictions with regard to architectural integration. The system presented in this article consists of a linear Fresnel reector which focuses radiation in a manner analogous to a lens. The receiver remains static and solar tracking is achieved in a simple and effective way by rotation of the individual reectors. Thus overall movement is minimised facilitating incorporation into buildings and offering different possibilities for suiting the varied requirements of specic installations. In order to compare and to easily identify advantages of the system proposed, in comparison with the concentrators mentioned above, three indicators are described: the Characteristic Length, the Transparency Coefcient and the Shading Coefcient. The Characteristic Length (CL) is dened as the quotient between the volume needed by the tracking system and the aperture area of the concentrator. When CL values obtained by the concentrator are low, building integration capabilities increase. In case of the design system, this parameter is highly decreased due to the rotational tracking, taking the same value than the mirror width. On the other hand 2-axis concentrators achieve CL levels equal to the square root of the concentrator area, practically disabling a proper integration. The Transparency Coefcient relates the transparent surface to the light with respect to the aperture area of the collector. If the concentrator surface blocks a big percentage or totally the incoming radiation, the number of integrating possibilities of the concentrating systems is very limited. Compact Linear Fresnel Reector systems and similar devices take low values of the CL indicator;

however the structure of the concentrator hinders both good views from the interior of the building and an adequate lightening. The inclination of the reectors (see Fig. 1) in the system described below facilitates to reach transparency coefcients guaranteeing comfort. The last of the indicators is the Shading Coefcient, which is the ratio between the direct beam radiation passing through the system and the total direct radiation received on the aperture area of the concentrator. Concerning to the 2-axis Fresnel concentrators, this coefcient exactly coincides with the Transparency Coefcient and it is equal to 0. In Compact Linear Fresnel Reectors these indicators are strongly dependent on the nal structure of the concentrating device. 2. System design and mathematical model The system design must aim to full the following requirements: architectural integrability (environmental integration, appropriate materials, dimensions that t the composition and harmony of the building, light weight), high compactness (this is the inverse of the aspect ratio, the aspect ratio being the ratio between the focal distance and the concentrator aperture), low mirror ratio (ratio between the surface area of the reectors and the concentrator aperture), high optical efciency and geometric concentration above 5X. The mathematical model is formed as follows: Let N be the total number of mirrors which form the concentrator, each being labelled with a subindex i. Cases with odd N are considered such that there exists a central reector with subindex i = 1 + (N 1)/2, although expressions described below can be extended for any case. The coordinate origin O is taken to be at the centre of rotation of the central mirror (Fig. 1). The axes of rotation of all reectors lie in the XZ plane and are parallel to the Z axis. The receiver, of width d, sits parallel to the XZ plane at a distance f from the origin. The length of the receiver is considered equal to that of the mirror. The distance of the centre of rotation of the ith mirror from the origin is dened to be xi, and the angle subtended at the centre of the receiver by the reector and the origin ai, where:

ai tan1

  xi f

If hs is the angle of incidence of the solar radiation (positive clockwise) then to reect the light onto the receiver each mirror must be inclined at:

Fig. 1. Schematic showing parameters used in the mathematical model.

D. Chemisana, J.I. Rosell / Energy Conversion and Management 52 (2011) 32413248

3243

bi

hs ai 2

As bi is directly proportional to hs, with a derivative term equal to @bi 1 ; it can be concluded that after the initial position of the mir@hs 2 rors has been established for a particular incidence, the relative movement required to achieve tracking is the same for all reectors. This permits the use of a single linear driver, representing an important mechanical advantage. Fig. 2 shows two phenomena which reduce the efciency of the system: shading and blocking. Shading refers to the interception of rays by a surface before incidence on the mirror under investigation and blocking to interception after reection from said mirror. In Fig. 3, a 3D view of the system proposed is shown to better appreciate the global appearance of the concentrating device. The position of the upper and lower limits of the shaded and blocked rays (respectively) have been calculated for each reector using an analytical ray tracing algorithm, involving the equations previously dened, programmed in Matlab. A diagram showing

the position of the end points of the mirrors and the shading and blocking limits, under an angle of solar incidence of 40, can be seen in Fig. 4. The reader should understand that rays that fall on the reector below the shading limit are shaded and those that fall above the blocking limit are blocked. These limits have been calculated using the algorithm which implements the projection and intersection of three lines on the XY plane, these being the line along which the mirror lies, the line followed by the extreme incident rays, and the line followed by the reection of these rays. Through this study we have also determined the size and position of the focal area on the receiver. The luminous band coming from each mirror has intensity (projected intensity, Ip) which depends on the angle of solar incidence and the inclination and position of the respective mirror. Id is the direct normal irradiance and qm is the specular reectivity of the mirrors.

Ip Id qm coshs bi cosai

Finally the irradiance prole produced at the receiver by each reector is calculated by taking the convolution:

Ii Ip i  Is i
Incident rays

1 1

Ip s Is x si ds

  2 exp 2x 2 is the effective intensity with rop p es f 2 x2 which the sun irradiates the receiver and rop cosbi i is the stanwhere Is x r
1 p op 2p

Blocking limit

Shading limit Reflected rays Receiver

Fig. 2. Schematic of shading and blocking in the system.

Fig. 3. 3D image of the system.

dard deviation which represents an ensemble of errors to which reective concentrators are subject in real situations, these arise from Gaussian-like sunshape, lack of perfect specularity, imperfections in reector placement and elevation, angular displacement of the receiver and tracking errors. The effect of all this errors is involved in the effective solar cone angle (es) [6]. An initial Base System (BS) has been dened which conforms to the geometrical constraints arising from the need to full the requirements outlined at the beginning of this section. In the BS, the mirror separation dm is equal to half the receiver width, the mirror width lm is the same for each reector and the aspect ratio Ar takes a value of 0.5. In a pre-optimisation stage, the mirror width was found to be the triple of the receiver width. Following denition of the BS, three other congurations which differ from the BS in one or more characteristics have been designed and analysed (see Fig. 5): Varied mirror Positions System (VPS): In this conguration the mirror separation dm is distinct for each pair of reectors and has been set such that, in the central position, the incident light is not shaded nor allowed to pass through the mirrors without reection: that is that the shading limit is exactly at the bottom end of each reector. The mirror width is kept constant. Varied mirror Widths System (VWS): Here, the mirror width lm is distinct for each reector and has been set such that, in the central position, the reected light occupies all the space between lower edges of the mirrors: that is that the blocking limit is exactly at the top of each mirror. The distance between the rotation axes and either edge of mirror is given an upper limit of 0.5 m so as not to produce an overly negative visual impact nor create mechanical difculties. The mirror separation is kept constant. Varied mirror Positions and Widths System (VPWS): Here, the width of each mirror has been set as in the previous conguration and the separations have been individually adjusted so as to maximise the transmission of light to the receiver. As the ray tracing technique is based on a geometric procedure, the system and its results are directly scalable. This permits to generalise conclusions to any size or scale by considering the dimensionless longitudes, i.e. by dividing all lengths by the aperture area of the concentrator.

Blocke d ray

3244

D. Chemisana, J.I. Rosell / Energy Conversion and Management 52 (2011) 32413248

y reference cordinate (m)

0.6 0.5 0.4 0.3 0.2 0.1 0 -1.5

Mirror top limit

Mirror bottom limit

Shading effect limit

Blocking effect limit

10 11 12 13 14 15 16 17 18 19 20 21

-1

-0.5

0.5

1.5

x reference cordinate (m)

Fig. 4. Point diagram of the mirrors in the Optimised Base System under 40 solar incidence.

PR PA

In order to clarify the differences between the different systems under analysis, two further variables have been incorporated: the Mirror Ratio Mr and the Mirror Efciency gm.
Fig. 5. Schematic of the different congurations.

Mr

Lm Lc Ip le I s lm

3. System optimisation 3.1. Objective function It can be seen in Fig. 6 that the systems optical efciency varies with solar position. To optimise the system we must dene an objective function which encapsulates all of this information. The objective function is the daily global optical efciency (gG), dened as the quotient of the solar energy received onto the receptor QR to the solar energy incident on the aperture QA over the course of a day where the irradiation conditions for the day are dened to be 6 h of uninhibited AM 1.5 illumination centred at solar midday and with an angular variation of solar incidence from 45 to 45 over this period. Only direct radiation is considered, diffuse and reected radiation are ignored.

gm

where Lm is the total length of the mirrors equal to the sum of the widths of the mirrors, Lc is the width of the concentrator aperture in the x direction (Fig. 1), le is the effective length of the mirror, i.e. the distance between the blocking and shading limits, and Ip, Is and lm are dened as in Section 2. 3.2. System discussion We proceed with an analysis of the different congurations outlined at the end of Section 2 based on the objective function and the ratios dened in Section 3.1. Each conguration has been optimised using a methodology of iterative calculations with the goal of determining which optimised system can maximise gG and for which choice of parameters it would do so. The BS and the VPS have been optimised by xing the mirror separations and inclinations to the values previously dened for each system and varying the width and position of mirrors along the line in the XY plane on which that mirror lies (see Fig. 1). The

gG

QR QA

The instantaneous optical efciency is dened as the quotient between the power arriving at the receiver PR and the power arriving at the aperture PA at a given instant.
100 90 80 70
OBS

OVPS

VPWS

VWS

(%)

60 50 40 30 20 0 10 20 30 40 50

Angle of incidence ()
Fig. 6. Instantaneous optical efciency of the systems as a function of the angle of solar incidence.

D. Chemisana, J.I. Rosell / Energy Conversion and Management 52 (2011) 32413248 Table 1 System efciency for normal solar incidence, daily global efciency and mirror ratio for the different congurations under investigation. System type BS OBS VPS OVPS VWS VPWS OVPWS Centred efciency (%) 66.78 73.17 88.25 92.8 96 50.5 78.77 Daily global efciency, gG (%) 51.87 56.38 53.63 56.2 43.32 53.11 56.6 Mirror ratio (Mr) 3.01 3.13 8.07 8.21 7.63 1.85 1.85 Comments

3245

model (4.1) and subsequently using a ray tracing numerical simulation (4.2).

4.1. Analytical results

Off-set mirrors Off-set mirrors Symmetric mirrors dOVPWS = 2 dBS

positions of the axes of rotation relative to the coordinate axes are xed. As with the initial denition of the VWS, the distance of the reector position may be varied above or below its axes of rotation and is set the upper limit of 0.5 m. The Optimised Base System (OBS) is achieved by moving the mirrors of the BS 14 cm above their centres of rotation whereas just as the Optimised Varied mirror Positions System (OVPS) is achieved by moving the mirrors centres of the VPS 24 cm above their centres of rotation. The central position efciency, global efciency and mirror ratio for the optimised systems are shown in Table 1. The global efciency of the VPWS can be improved upon by 3.5% if the receiver width is increased to be 4 times the BS one (Optimised Varied mirror Positions and Widths System OVPWS). This variation of the VPWS is not viable as the geometric concentration is reduced by half. The results are shown in Table 1. The analysis shows the VWS system has a much lower global efciency than the other systems; it is therefore considered the least viable option and not an optimised solution. It can be seen from Table 1 that the system which optimises the objective function whilst satisfying the critical parameters is the OBS. The OVPWS, despite having a higher global efciency (gG = 56.6%) is not a good alternative as the receiver is required to be 0.4 m width making the system impractical and reducing the geometric concentration to below 5X. Albeit OVPS and VWS have a very high centred efciencies, their mirror ratios are respectively 8.21 and 7.63, making them unsuitable.

4. Results In this section we present the principal results for the Optimised Base System (OBS) obtained rst by an analytical ray tracing

The following results are presented for the system which best satisfy requirements and considered the optimum, OBS, with N equal to 21 reectors. In the previously explained Fig. 4, it is shown the x and y coordinates of the end points of the mirrors in the OBS conguration for an angle of solar incidence of 40. The coordinates at which the lowermost unshaded ray and the uppermost unblocked ray impact each reector are also shown. It can be seen that for the rst 6 mirrors (left of the gure) none of the incident radiation is transmitted due to shading and blocking. Light is only transmitted by mirrors numbered 7 and above and full transmission to the receiver is only achieved by mirrors numbered 13 and above. The global efciency gG of the solar radiation on the concentrator aperture, compared to that of a at solar panel under the same conditions, varies in a distinct manner depending on the conguration which has been optimised (Table 1). It can be observed in Fig. 6 that for the OBS the prole is the most regular, having a maximum efciency of 73.17% at angle of 0 and a minimum of 54.0% for 36. Mirror efciencies of the OBS conguration, in function of solar angles and reectors position, are depicted in Fig. 7 and 8. It can be noted in Fig. 7 that the efciency peaks achieve a displacement in the direction of the sun movement. Efciency tendencies of the different Reectors, drawn in Fig. 8, change depending on where are positioned in the concentrator. Extreme mirrors reach minimum efciencies when the sun goes to their opposite side and central reectors get the best performance when the sun is centred with the system (hs = 0). In case of the extreme mirror number 1, the maximum efciency is equal to 72.08% at 45, 31.71% at 0 and drops to 0% when the angle of solar incidence is higher than 20. Fig. 9 shows the solar radiation prole across the width of the receiver for the OBS under a range of angles of solar incidence. The prole is symmetrical for normal incidence and becomes more skewed as the angle of incidence is increased (2 cm on the left for the highest). In all cases a small portion of the radiation falls outside of the half of the design receiver (10 cm). This fact arises the opportunity of increasing the average concentration ratio by decreasing the receiver width, considering then lower tracking tolerances. The highest power per unit area is almost 18,000 W/m2.

1.2 0 1 10 20 30 40 50

0.8

0.6

0.4

0.2

0 -1

-0.5

0.5

Reflector position (m)

Fig. 7. Optical efciency of each mirror (gm) as a function of reector position for a range of angles of solar incidence. Lines connect data points for different mirrors under illumination from the same incidence as a guide to the reader.

3246

D. Chemisana, J.I. Rosell / Energy Conversion and Management 52 (2011) 32413248

1.2

1
16 17 15

13

14 12

0.8

18 19 11 10 20 9 8 7 6 5 4 3 1 2

0.6

0.4

21

0.2

0 0 5 10 15 20 25 30 35 40 45 50

Angle of incidence ()
Fig. 8. Optical efciency of each reector as a function of the angle of solar incidence. Mirrors are numbered from the left of the device, the central mirror being number 11.

20000 0 18000 16000

10

10 20 30 40 50

Power per unit area (W/m )

14000 12000 10000 8000 6000 4000 2000 0 -0.1

20 30 40 0

50

-0.05

0.05

0.1

Position along the receiver width (m)

Fig. 9. Proles of irradiated power at the receiver for a range of angles of solar incidence. Normal incident irradiance is 1000 W/m2.

Finally, to point out that the Characteristic Length, the Transparency Coefcient and the Shading Coefcient for the Optimised Base System, when the sun is centred with respect to the optical axis, are found to be respectively 36 cm, 0.4 and 0.4.

4.2. Numerical ray tracing simulation and validation The optical behaviour of the Optimise Base System has been simulated using the commercial ray tracing software package OptiCAD in order to compare and validate the analytical results. The solar source was modelled in OptiCAD as a pillbox shape, with half angle set to 4.65 mrad. This means that all radiation coming from within the solar disc is assumed to be of equal intensity, and that no radiation originates from outside the solar disc [7]. OptiCAD simulations applied for optically characterise solar concentrators are widely used and experimentally validated. In the

specic eld of solar reectors, a great accuracy of the numerical results is reported by [8]. Solar tracking is performed by rotating the reectors until the locus of best foci is reached, i.e. when the maximum total power is received and when the prole is best centred on the receiver. Numerical results obtained in the simulations have been post processed to graph the 3D concentrating prole, referenced to the irradiance on the concentrator aperture (Fig. 10). By analyzing the colour contour lines, the concentration peak movement can be determined. Despite the peak displacement, the focal area width (around 20 times smaller than the concentrator aperture, covering the half of the receiver design width) and position over the receiver varies only slightly with the angle of incidence, as occurred in the analytical results (Fig. 9). As mentioned in Section 4.1, it is therefore possible to decrease the design width of the receiver in order to reach higher concentration ratios due to the narrower size of the focal area.

D. Chemisana, J.I. Rosell / Energy Conversion and Management 52 (2011) 32413248

3247

25
Concentration (suns) 2 6 10 14 18 22

25
Concentration (suns) 2 6 10 14 18 22

20

20

Concentration (suns)

15

Concentration (suns)

15

10

10

wi 10 dth 0 (cm )

50

Target width

(cm)

Target width (cm)

Fig. 10. Concentration prole at the receiver angle of solar incidences of 0 (left) and 45 (right). Data obtained using ray tracing software package OptiCAD.

Fig. 11 shows the comparative results between the analytical and numerical ray tracing analysis. It represents the average concentration over the receiver for two different widths according to the determined size of the focal area (dashed curves) and maximum concentration (solid curve) as a function of angle of solar incidence across the entire acceptance angle of the system (from 45 to 45). It can be seen that both average and maximum concentrations present maxima for extreme angle illumination and minima when the solar rays are incident at around 25. This behaviour stems from two causes: rstly the system is less subject to effects of shading and blocking when the system is oriented to receive illumination at an incidence of zero degrees, consequently when the illumination angle is deviated from this central position, the system efciency decreases. This phenomenon is compensated by the second cause which is that the conguration of the reectors causes the system to be much more effective and efcient at capturing rays with high angles of incidence than the at reference system used to dene the optical efciencies. Comparing the simulated results which those obtained through analytical algorithms, it can be seen that the behaviour of both models are almost identical for the mean concentration values
20

(for 10 and 20 cm receiver widths). Concerning the maximum concentration values, numerical results agree perfectly the analytical trends, although it has a more signicant variation of the values. This drawback specically and general divergences are explained due to the high sensitivity of this value to the exact position and orientation of the reectors, the statistical uncertainty of the numerical ray tracing and mainly because of two or more reections occurring and considered in Opticad but not taken into account in the analytical model. Analysing numerical results obtained in Figs. 10 and 11 in comparison with Fig. 9, and even though the relative efciency compared to the at reference system may increase as described in the analysis of Fig. 11, we can see in the absolute value of power ux shown in Fig 9 that on increasing the angle of incidence the energy capture diminishes. 5. Conclusions The transmissive Fresnel reector has an important potential for building integration applications due to three important characteristics:
30

Average Concentration (suns)

15 18

12 10 6

5 -45

0 -35 -25 -15 -5 5 15 25 35 45

Angle of incidence ()
Fig. 11. Analytical and numerical ray tracing representation of the average and maximum concentrations achieved by the OBS as a function of angle of solar incidence.

Maximum concentration (suns)

Numerical average concentration (receiver: 20 cm width) Analytical average concentration (receiver: 20 cm width) Numerical average concentration (receiver: 10 cm width) Analytical average concentration (receiver: 10 cm width) Numericall maximum concentration Analytical maximum concentration

24

Tar g

10 9 8 7 6 5 4 3 2 1

Ta 0 rg et

2019 18 17 16

15 14 13 12 11

20 19 18 17 16 15

14 13 12 11 10 9 8 7 6 5 4 3 2 1

et w

50

idth

100

(cm )

3248

D. Chemisana, J.I. Rosell / Energy Conversion and Management 52 (2011) 32413248

Reduced working volume related to the reduced value of the Characteristic Length for this conguration. The Characteristic Length is 36 cm for the Optimised Base System. High values of Transparency Coefcient and Shading Coefcient compared with single axis or double axis solar concentrators. In the system analysed both characteristics are disengaged because of the system shape changes during the solar tracking. The values achieved of both, the Transparency Coefcient and the Shading Coefcient, lead to a useful illumination control of the building interior space. On analysing of different Fresnel transmission systems the following conclusions have been formed: It is possible to control the Fresnel transmission system using a single driver greatly improving technical practicability and reducing cost. In all studied cases the system efciency depends on the angle of solar incidence. The instantaneous efciency for elevated angles of incidence can be higher than that for normal incidences, an effect which is owed exclusively to the change in geometric structure of the concentrator at different times of day. The Mirror Ratio is here dened as the reector surface required per unit area of system aperture and permits evaluation of economic viability of the studied device. Based on this criterion, VWS and OVPS systems have been discarded (7.63 and 8.21 respectively), despite reaching very high efciencies (above 90%) when centred with respect to the Sun. Numerical ray tracing simulation agree and t the analytical results with a great concordance, albeit in the maximum concentration ratios are shown a few divergences due to uncertainty inherent to the numerical data treatment, number of mirror reections and accuracy in positioning and tilting of the reectors. The daily global efciency of the concentrator is an adequate parameter for determination of the validity of a system for a determined application. The Optimised Base System (OBS) is the studied conguration which best adequate the daily global optical efciency, reaching a value of 56.38%. The maximum concentration ra-

tio achieved by the OBS is predicted to be 18 suns at an angle of solar incidence of 0. The illumination prole across the receiver width has also been obtained using the above mentioned methods. The non-uniformity of the illumination prole is seen to increase with increasing angle of incidence. It is observed a displacement of the prole maximum with the sun movement from the centre of the receiver but its value is lower than 2 cm for the highest angle of incidence. Concerning that the ray tracing technique is based on a geometric procedure, the system and its results are directly scalable. This permits to generalise results to any size or scale by considering the dimensionless longitudes, i.e. by dividing all lengths by the width of the aperture area of the concentrator. Acknowledgement The authors would like to acknowledge funding by the Ministry of Science and Technology MCYT (ENE2007-65410). References
[1] Chemisana D. Building integrated concentrating photovoltaics: a review. Renew Sustain Energy Rev 2011;15:60311. [2] Tripanagnostopoulos Y, Siabekou Ch, Tonui JK. The Fresnel lens concept for solar control of buildings. Sol Energy 2007;81:66175. [3] Mills DR. Advances in solar thermal electricity technology. Sol Energy 2004; 76:1931. [4] Rosell JI, Vallverdu X, Lechon MA, Ibez M. Design and simulation of a low concentrating photovoltaic/thermal system. Energy Convers Manage 2005; 46:30343046,. [5] Vasylyev S. Nonimaging reective lens concentrator. In: Proceedings of the international conference on solar concentrators for the generation of electricity or hydrogen; 2005. [6] Rabl A. Active solar collectors and their applications. New York: Oxford University Press; 1985. [7] OPTICAD Optical Analysis Program, Opticad Corporation, 511 Juniper Drive, Santa Fe (NM), 87501 USA. [8] Schiricke B, Pitz-Paal R, Lupfert E, Pottler K, Pfander M, Riffelmann K-J, et al. Experimental verication of optical modelling of parabolic trough collectors by ux measurement. J Sol Energy Eng 2009;131:10049.