OSA's Digital Library

Energy Express

Energy Express

  • Editor: Christian Seassal
  • Vol. 21, Iss. S4 — Jul. 1, 2013
  • pp: A642–A655
« Show journal navigation

Illumination performance and energy saving of a solar fiber optic lighting system

David Lingfors and Tarja Volotinen  »View Author Affiliations


Optics Express, Vol. 21, Issue S4, pp. A642-A655 (2013)
http://dx.doi.org/10.1364/OE.21.00A642


View Full Text Article

Acrobat PDF (1793 KB)





Browse Journals / Lookup Meetings

Browse by Journal and Year


   


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools

Share
Citations

Abstract

The illumination performance and energy savings of a solar fiber optic lighting system have been verified in a study hall - corridor interior. The system provides intensive white light with a high luminous flux of 4500 lm under 130000 lx direct sun radiation at a 10 m fiber distance from the sun-tracking light collector. The color temperature that describes the light color perceived is 5800 ± 300 K, i.e. close to the direct sunlight outside, and the color rendering index (86), that describes how well colors are rendered under the light source, is higher for the solar lights than for the supplementary fluorescent lights (77). Thus this high quality solar lighting improves the visibility of all kinds of objects compared to the fluorescent lights. Annual lighting energy savings of 19% in Uppsala, Sweden and 46% in southern Europe were estimated for a study hall interior, as well as 27% and 55% respectively in an interior illuminated 16 h per day all days of a year.

© 2013 OSA

1. Introduction

A system that captures the high intensity direct component of solar light, focuses it by a concentration of the order of 2000 into an optical fiber and distributes the visible part of it into buildings would be an ideal lighting supplement to artificial lights in commercial buildings. Such a system provides high quality light; shows excellent illumination performance and saves much more lighting energy than low-energy artificial lights, dependent on the weather and geographic location, as shown in this study.

This study is not the very first invention of this technology [1

1. T. Nakamura, “Optical waveguide system for solar power applications in space,” Proc. SPIE 7423, 74230C, 74230C-10 (2009). [CrossRef]

4

4. Himawari solar fiber optic lighting systems, http://www.himawari-net.co.jp/e_page-index01.html.

], but up to now the potential lighting performance and energy savings of this technology have been unknown. In addition, partially new characterization methods were developed in this project and are here reported for first time. This report thus aims to thoroughly verify the excellent lighting performance and energy savings of a newly launched solar fiber optic lighting technique and to describe the characterization methods used for this purpose.

Solar fiber optic lighting systems have been available for the past 20 years [4

4. Himawari solar fiber optic lighting systems, http://www.himawari-net.co.jp/e_page-index01.html.

], and have been studied in several reports [1

1. T. Nakamura, “Optical waveguide system for solar power applications in space,” Proc. SPIE 7423, 74230C, 74230C-10 (2009). [CrossRef]

3

3. F. Francini, D. Fontani, D. Jafrancesco, L. Mercatelli, and P. Sansoni, “Solar internal lighting using optical collectors and fibres,” Proc. SPIE 6338, 63380O, 63380O-8 (2006). [CrossRef]

,5

5. T. T. Volotinen, N. Nilsson, D. Johansson, J. Widen, and Ph. Kräuchi, “Solar fibre optic lights -daylight to office desks and corridors,” Proceedings CISBAT2011, 491–496 (2011).

8

8. M. S. Mayhoub and D. J. Carter, “Hybrid Lighting systems: performance and design,” Lighting Res. Tech. 44(3), 261–276 (2012). [CrossRef]

]. Breakthroughs of this technology have been, however, delayed due to technical problems [6

6. L. C. Maxey, M. V. Lapsa, P. Boudreaux, D. D. Earl, J. Morris, and T. Bunch, “Hybrid Solar lighting: Final technical report and results of the field trial program”, U.S. Department of Energy (DOE) Information Bridge, http://www.osti.gov/bridge, ONRL-TM-150, (2008).

], a too high price [4

4. Himawari solar fiber optic lighting systems, http://www.himawari-net.co.jp/e_page-index01.html.

,8

8. M. S. Mayhoub and D. J. Carter, “Hybrid Lighting systems: performance and design,” Lighting Res. Tech. 44(3), 261–276 (2012). [CrossRef]

10

10. A. Kribus, O. Zik, and J. Karni, “Optical fibers and solar power generation,” Sol. Energy 68(5), 405–416 (2000). [CrossRef]

], a low light output power [1

1. T. Nakamura, “Optical waveguide system for solar power applications in space,” Proc. SPIE 7423, 74230C, 74230C-10 (2009). [CrossRef]

3

3. F. Francini, D. Fontani, D. Jafrancesco, L. Mercatelli, and P. Sansoni, “Solar internal lighting using optical collectors and fibres,” Proc. SPIE 6338, 63380O, 63380O-8 (2006). [CrossRef]

,5

5. T. T. Volotinen, N. Nilsson, D. Johansson, J. Widen, and Ph. Kräuchi, “Solar fibre optic lights -daylight to office desks and corridors,” Proceedings CISBAT2011, 491–496 (2011).

9

9. The earlier version (SP2) of Parans Solar lighting system, www.parans.com.

] and fairly low general knowledge of this technology. Comparisons have been reported with so called “skylights”, tube reflectors and daylight providing transparent building structures (roof windows, glass walls and atrium “glass castles” in the middle of buildings) [6

6. L. C. Maxey, M. V. Lapsa, P. Boudreaux, D. D. Earl, J. Morris, and T. Bunch, “Hybrid Solar lighting: Final technical report and results of the field trial program”, U.S. Department of Energy (DOE) Information Bridge, http://www.osti.gov/bridge, ONRL-TM-150, (2008).

8

8. M. S. Mayhoub and D. J. Carter, “Hybrid Lighting systems: performance and design,” Lighting Res. Tech. 44(3), 261–276 (2012). [CrossRef]

]. The comparisons that consider the solar fiber optic light system as an ordinary daylight system by using a static (not sun-tracking) horizontal or vertical pyranometer reading as a measure for the sun radiation [7

7. M. S. Mayhoub and D. J. Carter, “Towards hybrid lighting systems: A review,” Lighting Res. Tech. 42(1), 51–71 (2010). [CrossRef]

,8

8. M. S. Mayhoub and D. J. Carter, “Hybrid Lighting systems: performance and design,” Lighting Res. Tech. 44(3), 261–276 (2012). [CrossRef]

], have not provided the full picture of the lighting potential of this technology. The output intensity of the solar light system is not linearly dependent on the overall daylight [8

8. M. S. Mayhoub and D. J. Carter, “Hybrid Lighting systems: performance and design,” Lighting Res. Tech. 44(3), 261–276 (2012). [CrossRef]

]. Instead, it is important to realize that a fiber optic solar light collector of the system studied here is an effective sun-tracker that couples the direct, highest intensity component of the sun light, not the overall diffuse daylight, into the fibers. Here we show a linear dependency of the output luminous flux of the SP3 system as a function of the direct intensity of solar light. Another important aspect that has not been fully reported earlier [1

1. T. Nakamura, “Optical waveguide system for solar power applications in space,” Proc. SPIE 7423, 74230C, 74230C-10 (2009). [CrossRef]

9

9. The earlier version (SP2) of Parans Solar lighting system, www.parans.com.

] is the spectrum of this solar light, and the benefits for human beings only such a light can provide. The third significant property is the energy saving that is provided.

Daylight consists of the direct radiation and diffusely scattered components. During clear sky conditions, the direct visible solar illuminance is of the order of 100 000 – 140 000 lx in Sweden, while the diffuse solar light is well below 10 000 lx. When clouds pass by and hide the sun, the direct visible component decreases down to 20 000 – 50 000 lx, while the diffuse radiation increases. During cloudy weather the sunlight illuminance is below 10 000 lx to all directions, i.e. all daylight is diffusely scattered.

In optics and photonics an input or output power of light is measured in watts. By definition the monochromatic light power 0.1464 mW at 555 nm corresponds to the luminous flux of 1 lm. Thus at 555 nm the illuminance of 1 lx corresponds to a 0.1464 mW/m2 optical power per square meter area, and a direct solar illuminance100 000 lx ~14.6 W/m2 [11

11. CIE 84–1989 Photocopy Edition 1996, “The measurement of luminous flux”, Technical report, CIE, Austria (1996).

14

14. C. S. McCamy, “Correlated color temperature as an explicit function of chromaticity coordinates,” Color Res. Appl. 17(2), 142–144 (1992). [CrossRef]

].

New, low-energy fluorescent lights and light emitting diodes (LED) are used in daytime for lighting of commercial buildings where lighting stands for 25 - 45% of the energy consumption [15

15. M.-C. Dubois and Å. Blomsterberg, Energy saving potential and strategies for electric lighting in future North European, low energy office buildings: A literature review. Energy and Buildings (2011), doi: [CrossRef] .

]. However, it has recently been reported that there is a trade-off between the efficacy and light quality for these sources [16

16. Y. Yang and X. Shi, “Analysis and evaluation on luminaire efficacy and colour quality of LED downlights”, Proc. of CIE2012 Lighting quality and energy efficiency-conference CIE x037:2012, pp 05, 546–554, (2012).

]. The efficacy for a light source is the ratio: the total luminous flux in lumens divided by the consumed electrical power in watts. – Thus those light sources that provide the highest energy efficacy provide only a moderate quality of light (low color rendering index (Ra), non-optimum correlated color temperature (Tcp), low light fidelity etc.) [12

12. CIE 15:2004 3rd Edition, “Colorimetry”, Technical report, CIE, Austria (2004).

14

14. C. S. McCamy, “Correlated color temperature as an explicit function of chromaticity coordinates,” Color Res. Appl. 17(2), 142–144 (1992). [CrossRef]

,16

16. Y. Yang and X. Shi, “Analysis and evaluation on luminaire efficacy and colour quality of LED downlights”, Proc. of CIE2012 Lighting quality and energy efficiency-conference CIE x037:2012, pp 05, 546–554, (2012).

]. Those light sources, e.g. LED lamps that consist of a selection of LED:s of different spectral bands, which together emit a sum spectrum close to the sun light (high Ra, high Tcp, high light fidelity, full spectrum) have only a moderate or low efficacy [16

16. Y. Yang and X. Shi, “Analysis and evaluation on luminaire efficacy and colour quality of LED downlights”, Proc. of CIE2012 Lighting quality and energy efficiency-conference CIE x037:2012, pp 05, 546–554, (2012).

].

The new energy saving targets for European buildings [15

15. M.-C. Dubois and Å. Blomsterberg, Energy saving potential and strategies for electric lighting in future North European, low energy office buildings: A literature review. Energy and Buildings (2011), doi: [CrossRef] .

], as well as the requirements defined by the Green certificates for buildings [15

15. M.-C. Dubois and Å. Blomsterberg, Energy saving potential and strategies for electric lighting in future North European, low energy office buildings: A literature review. Energy and Buildings (2011), doi: [CrossRef] .

], mean that most of the energy consuming technologies in buildings must be further improved. The energy consumption of lighting has been already decreased from the level it was 20 years ago by about 60% in many large commercial buildings by replacing the old fashioned electric lights by the least consuming T5-technology (105 lm/W efficacy) and by automatic regulation of the lighting usage time. However, there remains much more to improve, and at the same time, the quality of the light could also be improved by maximizing the use of solar light by the fiber optic solar lighting technology examined here.

The energy saving LED technologies are on the commercial market [17]. The LED lighting systems are available with the efficacy ranging from 30 to 120 lm/W [17]. Solid state light (SSL) sources and organic light emitting diodes (OLED) lighting technologies are still in progress of development, but are estimated to have the same order of electric efficacy as the LEDs. However, the solar lighting technology with its significantly greater efficacy and excellent light quality, as shown in this paper, can replace electric lights completely during sunny weather, and thus would increase the energy saving of any supplementary artificial lighting, dependent on the time when direct sun radiation is available.

The light quality and purchase price also matter, and require much more attention than has been given until now. For example for an office, several thousands of EURs are normally used per person to provide a good quality, highly ergonomic chair, desk and computer. But the lighting equipment and installation costs and especially the light quality are neglected, or even worse, are purposely kept as a so low portion of the total office facility costs as ever possible. Furthermore in European countries, if someone gets a pain in his/her back due to non-ergonomic office furniture or working conditions, he or she will get free of charge expert help to solve those problems. - The possible problems with light quality are not recognized or known by most people. No expertise or help is available for someone working under a low quality light, which can cause problems: such as tiredness, depression, low visibility of details and color recognizing problems, non-balanced melatonin production due to a too low or too high proportion of blue light, low working productivity, etc [18

18. L. Edwards and P. Torcellini, A literature review of the effects of natural light on building occupants”, National Renewable Energy Laboratory, Colorado, USA, NREL/TP-550–30769 (2002).

].

2. Experiments

2.1. The test site and equipment

The fine adjustment of the collector direction is performed by a light sensor directed in parallel with the lenses on the panel, located in the lowest end on the right black stripe seen across the panel. For the tests reported here, and for the optimization of the energy savings and the control signal for the dimming of the artificial lights, an extra sensor (Kipp & Zonen SP Lite 2 silicon pyranometer) has been attached to the solar panel, as is seen on the upper end of the left black stripe of the panel in Fig. 1(a).

The light collector of the SP3 system contains 36 square-shaped Fresnel lenses (65 mm x 65 mm, Fig. 2
Fig. 2 (a) Schematic description of the sun light collector and fibre luminaire of the SP3 system. (b) The fiber luminaire without the Spot luminaire L3, shown in picture (c). Most of the tests were performed without the spot luminaires.
). The lenses focus the sunlight onto the optical fiber ends. UV- and NIR-radiation reflecting filters, placed on the fiber ends, protect the fibers from the harmful components of the solar radiation. The 36 plastic optical fibers are bunched together into six cables that transport the light over a 10 m distance into the fiber luminaires installed at the ceiling of the test site: a corridor and study hall area (Fig. 1(b)). High attenuation of the visible light (2 dB per 10 m length) in the fluorinated polymethylmetacrylate (PMMA) material of these fibers limits the light transport distance to about 20 m.

The lighting conditions in the test room have been simulated by the DIALux-CAD software suitable for planning the indoor lightning, by taking into account the electric lights, effects of the doors, floor, ceiling and window (Figs. 3-5). The wall is painted by glossy white, which has an estimated reflection factor of 0.70. The ceiling consists of square 60 cm x 60 cm, white panels with matt grey-white texture. The reflection factors were estimated to 0.70 for the ceiling and 0.35 for the linoleum floor. The window is a double glass window of model Schott Pyran S® 6 mm/glass with a visual transmission factor of 0.81. However, the glass surface is coated which gives a lower transmission, estimated to 0.75. Two types (Fagerhult and Focus Lighting) of fluorescent down light luminaires are used at the test site and the technical data and light distribution are shown in Fig. 4
Fig. 4 Information of the illumination properties of the fluorescent lights at the test site.
. Both are high frequency luminaires.

2.2. The measurements of illuminance and luminous flux

The direct visible illuminance of the sun, and the illuminance distributions over the test site area have been measured with a Hagner EX-4 lux meter equipped with a Vλ- wavelength filtered and cosine-direction corrected SD2 detector. The illumination distributions have been measured for different weather conditions and lighting arrangements at the 85 cm distance from the floor at the network grid points shown in Fig. 3, according to the CIE 84-1989 standard11.

The luminous flux (total light output intensity) for each fiber luminaire has been measured repeatedly under varied sun intensity conditions with the same lux meter as the illuminance, by using it connected to an integrating sphere of 15 cm in diameter. The sphere was calibrated together with the lux meter at the SP Technical Research Institute of Sweden in Borås with a standard light source by pointing the light into the sphere and using a fiber bundle similar to the SP3 fiber luminaires between the standard light source and the sphere. This gave a conversion factor (1/114) from the lux values to the luminous flux values for the fiber luminaires.

In order to define the light color and to understand the quality of the light, the spectral distribution over the wavelength range 300 – 1000 nm has been measured for the sun, fiber luminaires and fluorescent lights by the SI-OSMA Optical spectrum analyzer, equipped with a Jarrel Ash monochromator JA-150 and a Princeton Instrument detector: Eiry1024- SI-IRY 1024/L. Neutral density filters of 3.0 - 7.0 have been used because of the high sensitivity of the detector, and the high intensity of the direct solar light. The correlated color temperatures (Tc) have been calculated according to the McCamy method [12

12. CIE 15:2004 3rd Edition, “Colorimetry”, Technical report, CIE, Austria (2004).

14

14. C. S. McCamy, “Correlated color temperature as an explicit function of chromaticity coordinates,” Color Res. Appl. 17(2), 142–144 (1992). [CrossRef]

] and the color rendering indices (Ra) were measured according to the CIE 15:2004 [12

12. CIE 15:2004 3rd Edition, “Colorimetry”, Technical report, CIE, Austria (2004).

] and calculated with the CIE 013-3 [13

13. CIE13, 3–1995, “Method of measuring and specifying colour rendering properties of light sources”, Technical report, CIE, Austria (1995).

] standard software.

3. The results

3.1. Illumination in the test area

The illuminance distributions, measured at the marked grid points (Fig. 3), are shown in Fig. 6
Fig. 6 The light distribution at the test site area, measured during four different weather and lighting conditions. (a) Cloudy weather, only emergency lights are on, (b) very cloudy day, the electric lights are 100% on. (c) Sunny weather, the electric lights are 100% on. (d) Sunny weather, the electric lights are dimmed down to 3% level, and the solar fiber optic lights, window and emergency lights are providing the light.
. The graph (a) shows the background illuminance distribution that comes from the window, glass doors and the emergency light close to the elevator shafts under cloudy weather. Graph (b) shows the full illuminance from the electric lights during a very cloudy day. Graph (c) shows full illuminance during a sunny day, when the electric lights are 100% on. The graph (d) shows the illuminance during a sunny day, with the artificial lights dimmed down to 3% and the solar fiber lights providing the full illumination. The area with over 300 lux illumination (Fig. 6(d)) is significantly greater with the solar fiber optic lights than with the fluorescent lights on (Fig. 6(c)).

An area over 500 lx is also obtained (Fig. 6(d)) far away from the window during the sunny weather, even though the electric lights dimmed down to 3%. The illuminance results correspond to the simulated behavior in DIALux (Fig. 5). The obtained light distributions from the solar lights were, however, slightly wider (Chap. 3.3) than the DIALux simulation data indicated for them.

3.2. The luminous flux and efficacy

Average luminous fluxes of 767 lm and 430 lm per luminaire and total output fluxes of 4600 and 2580 lm respectively, were obtained during sunny weather of 130 000 lx and 98 000 lx direct sun illuminance (Fig. 7
Fig. 7 The measured output luminous fluxes of the fiber luminaires of the Parans SP3 solar fiber optic lighting system at 10 m and 20 m fiber lengths under various direct sun illuminance.
) at 10 m fiber distance from the sun collector. At 20 m fiber distance 400 – 500 lm per luminaire were obtained under 100 000 – 130 000 lx.

The electric power consumed by the SP3 system is 10 W during daytime and 1.8 W in the stand-by mode at night, implying the lighting efficacy 460 lm/W at a 10 m fiber distance when the direct sun illuminance is 130 000 lx and 260 lm/W at 98 000 lx, thus being about 300 lm/W in average during the sunny weather. The corresponding efficacies at 20 m fiber distances are 330 lm/W and 200 lm/W, which still are well above all commercially available electric light sources.

The luminous fluxes of the fiber luminaires and also the efficacy of the system are linearly dependent on the intensity of the direct visible sun irradiance down to about 65 000 lx, according to these results. The output flux and efficacy goes down to about half when the sun intensity decreases from 130 000 lx to 65 000 lx (Fig. 7). The obtained efficacies are much higher than for commercially available LED-lamps (30 – 120 lm/W) and for the most energy efficient fluorescent lights (type T5, efficacy 105 lm/W), and significantly higher than the efficacies of the fluorescent lamps of the test site (30 – 65 lm/W, Fig. 4).

The light output efficiencies were calculated from the total direct sunlight luminous flux at the surface area of the Fresnel lenses of the sun collector (being 4200 lm/lens at 100 000 lux direct sun illuminance) compared with the measured luminous flux output (430 lm) for the 6 fiber luminaires at the 10 m and 20 m fiber distances. The output efficiencies were thus only 19% and 17% at these fiber distances respectively. The coupling loss of this system is estimated to be about 50%. The rest of the loss is caused by the attenuation inside the plastic fibers (ca. 2 dB/10 m).

3.3. The light distribution

The light distribution was measured for each type of luminaire measured in the test site at 85 cm height from the floor. The nadir intensity is the intensity measured in the axial direction of the fiber, i.e. directly under the fiber luminaire. The results, from which the background illumination has been subtracted (Fig. 8
Fig. 8 The DIALux manufacturer information of the light distributions and the measured distributions for a solar fiber optic luminaire under 130 000 lx sun illuminance (a) without the L3 spot diffusor and (b) with the L3 spot diffusor, and for the fluorescent lights, (c) Focus Lighting 22022 and (d) Fagerhult 76240 of the tests site.
), show that the nadir illuminance over 500 lx is obtained under the bare solar fiber luminaires (Fig. 2(a)) without L3 spot diffusers (Fig. 2(b)) in the test area, at 130 000 lx direct sunshine and even at lower sun illuminance down to 95 000 lx. With the spot L3 diffuser the nadir intensity over 4700 lx was obtained, which is much higher than what can be considered comfortable (below 2000 lx). Thus the spot diffusors were not used for any other tests.

The illuminated total area would decrease from about 13 m2 to about 6 m2 as the direct sun illuminance decreases from 130 000 lx to 65 000 lx (Fig. 8 and Table 1

Table 1. Illuminated area of the solar luminaires with and without the L3 spot and the two artificial lights at the test site at three illuminance levels, estimated from the data of Fig. 8.

table-icon
View This Table
| View All Tables
), if the fluorescent lights were shut off. The two types of old fluorescent lights showed at maximum 40 lux and 220 lux at nadir respectively (Figs. 8(c) and 8(d)), but they are not operating at their highest possible output in order to elongate their lifetime.

It was also found out that the fiber luminaires provide a wider light distribution than is estimated by the DIALux files available for SP3 system luminaires (www.parans.com) as shown in Figs. 5(b) and 8(a). Furthermore, due to the high total flux of these luminaires, the light could be distributed over a wider area by using efficient light diffusers. About a 5-8 m2 area could be illuminated by each luminaire at 100 lux illuminance (the required illuminance for a corridor type of interior). It would mean altogether 26 – 45 m2 illuminated area for one SP3 system of six fiber luminaires during 100 000-130 000 lx direct sun illuminance.

3.4. The spectrum and color of the light

The measured, normalized intensity spectra of the light received from the solar fiber luminaires, fluorescent lights and sun outside are shown in Fig. 9
Fig. 9 (a) The measured spectra of the solar light obtained from the fiber luminaires at nadir and at 15° angle under a 125 000 lx direct solar illuminance and for the Fagerhult fluorescent lights of the test site. (b) The spectra for 10 m and 20 m fiber distance from the solar collector. (c) The spectrum of the sun, measured under the same conditions as the spectra of solar luminaires (a) - (b).
. At the direct sun illuminance over 50 000 lx, i.e. when a significant output flux is available from the solar fiber luminaires, no significant spectrum or color changes of the light were obtained as a function of day time or weather. (Figs. 9(a)-9(b)). The color temperature was 5800 ± 300 K, and the color rendering index 84.9 ± 0.5 for the light from the 10 m system. These parameters are close to the direct sun color temperature values 6000 ± 300 K and only slightly lower than its color rendering index (98 ± 1).

The spectrum obtained from the fiber luminaires consists of almost all visible wavelengths. In contrast, the spectrum of the fluorescent lights consists of a few narrow, high intensity peaks at a few wavelengths as shown in Fig. 9, while other wavelengths between them are of low intensity or completely missing. The spectrum of the solar luminaires does not show any significant changes as a function of the distribution angle (Fig. 5(a)), which means that the light is perceived to have the same color in all directions. However, the spectrum for the 20 m fiber length shows deeper absorption dips at 540, 630 and 730 nm than the 10 m fiber luminaire due to the higher absorption in the plastic fiber material. Thus the color of the light is dependent on the fiber length for the kind of fibers used for this system.

The sun spectrum was studied by using a 2 m fiber bundle as a measurement probe and also without it. The direct sun spectrum was obtained to vary only a little with the time of the day, when the direct intensity was high enough for the solar lighting system, i.e. above 40 000 lx. The correlated color temperature was also found to depend only slightly on the weather. During cloudy weather very high Tc in the order of 7000 K was obtained for the sun light, while during clear sky in the direction of the sun, color temperatures of 5800- 6500 K were obtained. A typical direct sun light spectrum measured at the same day at the same time as the spectra of the fiber luminaires is shown in Fig. 9(c). The spectrum was found to depend on the direction of the probe, and therefore a laboratory sun-tracker was used to be certain to get the direct spectrum relevant for the solar lighting system studied. The direct sun intensity was found so high that neutral density filters (of optical density 3.0 – 7.0) were needed. The short optical fiber probe was found to be a practical tool for the optical spectrum analyzer to get the repeatable spectrum data of the direct sun radiation.

The results (Fig. 9 and Table 2

Table 2. The correlated color temperature (Tc), and color rendering index (Ra) of the light sources in Figs. 9(a)9(c).

table-icon
View This Table
| View All Tables
) revealed that the solar luminaire light contains the blue end of the sun spectrum as it is. The green and yellow wavelengths come out also very well, and even the red light is obtained. Thus the solar light system provides a fuller spectrum light than the fluorescent lamps. This can explain why this light has been experienced as a more natural, white and intensive light, under which all kinds of objects, their colors and structural details can be seen with a higher visibility than under the fluorescent lights at our test site.

4. Control signal for the light balancing

Solar fiber optic lights need a supplement from artificial lights in all kinds of interiors. Thus a control signal is required for the dimming of the artificial lights to balance the illuminance. The control signal shall reliably show the need for the additional light momentarily, and the automatically controlled electric lights need to respond quickly enough without delays to the signal. Several alternative arrangements were tested and a sufficient control signal was found in this study. An extra pyranometer, attached on the top of the solar light panel (Fig. 1(a)) and thus continuously and accurately tracking the sun, was found to provide a suitable and reliable solution. It showed a signal that linearly corresponded to the light intensity coupled into the fibers in the solar collector, and thus it also linearly correlated to the output luminous flux in the interior illuminated by the system. The output signal of the pyranometer was found to show a signal that linearly correlated to the signal of an illuminance sensor (a SD2 detector of Hagner EX-4 lux meter), when both were attached at the solar collector of the SP3 system and both pointed accurately towards the sun, tracking it continuously. The output signal of the pyranometer was thus used to show the direct sun illuminance (Fig. 10
Fig. 10 The measured luminous flux of one solar fiber luminaire as a function of the direct sun illuminance measured at the solar collector, during a partly cloudy day with quickly passing clouds.
)

A tube that would effectively prevent the diffuse sunlight from reaching the pyranometer, as is used in solar radiation analyses, was tested, but it was found problematic. The wind and a heavy bird, who liked to sit on the tube during the nights and mornings, caused more problems than could be withstood. The limiting of the optical aperture of the control signal sensor is, of course, reasonable, but it should be performed in an optical way, not by an out-sticking tube that can be moved by wind and birds and thus will disturb the tracking function of the light collector.

The luminous flux from a fiber luminaire in the test room was measured as a function of the direct illuminance at the pyranometer and showed an almost linear response on a day with quickly passing clouds (Fig. 10). The slight variation of the data depends on the 1-9 s delay in the measurement light output flux data caused by the 10 s interval and response time of the web logger. The response time of the pyranometer is below 1 s. The Si-detector that continuously measures the intensity and tracks the direct sun intensity on the system panel and which would be directly connected to the dimming driver of the artificial lights (without any web logger), would have the shortest possible response time, and would thus provide a higher correlation between the direct sun intensity and the luminous flux output. The delay in the response of the artificial lights could be minimized.

Interestingly, the passing clouds have been seen to cause a typical speed for a decrease of the direct sun intensity from 110 000 lx to 40 000 lx in about 5 seconds. It is known that human eyes are adapted to these smooth lighting changes quite well outside and when these changes occur behind a window. But in the indoor test location the fastest changes of the illuminance level and the 3 – 9 s delayed response of the artificial lights, have been perceived as disturbing, which implies the importance of a fast and reliable light balancing.

The output of the pyranometer was used as a control signal for the electric lights and tested at the test site for several weeks. It was found to work properly for most of the lighting conditions. Only very windy weather and quickly passing clouds, which both cause a variation of the direct sun intensity and an inaccuracy in the sun-tracking of the solar light system, has been considered less comfortable, causing noticeable illuminance variations due to a slightly too long delay in the light balancing.

5. Energy savings of the solar light system

In a study hall area like the test site, the artificial lights are turned on at 07.00-20.00 o’clock on holiday-free weekdays, which is approximately 250 days/year. This time is denoted as “study hours” from here on, thus the energy saving has been only recorded during the study hours. The total study hours time tsh is 3250 h.

Figure 11
Fig. 11 The sunrise and sunset in Uppsala 2012. The yellow area (3954 h) represents the daylight hours between 07:00 and 20:00 o’clock. The pink area represents the total daylight hours of the year (4506 h).
shows the sunrise and sunset times in Uppsala over the year 2012. The yellow area represents the total amount of the hours (3954 h) between 07:00 and 20:00 o’clock that coincide with daytime. The pink area represents the total daylight hours of the year, which is 4506 h in Uppsala. There are 1790 hours of sunshine in average per year in Uppsala [19

19. Swedish Meteorological and Hydrological Institute (SMHI), http://www.smhi.se/klimatdata/meteorologi /stralning/normal-solskenstid-for-ett-ar-1.3052, 2011.

]. Considering that a year in average has 365.25 days, the study hours of sunshine ts can be calculated as:

ts=250365.25395445061790h=1075h.
(1)

The sunlight system starts to track the sun 30 minutes before sunrise and stops 40 minutes after sunrise, equal to 1.17 hours in total. This gives the active time, ta, of the SP3 system for which the power consumption (Pa = 10 W) is higher than during the passive time (Pp = 1.8 W). The active time ta is thus:

ta=3954h/y+1.17h/day365.25day/y=4380h/y.
(2)

The annual passive time tp is also (8760 h – ta) = 4380 h. The time that the artificial lights would be on without the solar light system is equal to the total study hours. The power consumed by the fluorescent lights is Pf. Annual power consumptions without the SP3 system P1 and combined solar and electric light system at the test site P2 can be calculated as:

P1=Pftsh,
(3)
P2=Pata+Pptp+Pdts+Pf(tshts).
(4)

The total energy consumption of the solar light system is 51.7 kWh per year. The electric lights of the test site would consume at the 3% dimmed down state 525.6 kWh per year, and at 100% 2260 kWh per year. The power Pd consumed by the dimmed electric lights at 3% level was measured to be 60 W for the entire test area. The power for the electric lights Pf at 100% was 258 W. Thus the total annual saving of the lighting energy in Uppsala would be only 19%, but in a southern European location, e.g. in Italy with 3400 yearly sunny hours would result in 46% saving of the lighting energy in a similar study hall area.

A higher saving of the lighting energy would be achieved in such an interior where the lights are used every day and where most sunny weather would occur during the service time and thus benefit the saving. The saving would be 27% in Uppsala and 55% in Italy in such an interior of daily 16 h usage time. Furthermore, the electric lights could be switched off during the sunny weather instead of dimming. The highest lighting energy saving, 42% in Uppsala, would be reached in offices of 9 h daily usage time. One SP3 system could illuminate six offices during sunny weather. The lighting energy of 108 W per occupied office is assumed.

6. Conclusions

The lighting performance of a solar fiber optic lighting system has been verified by luminous flux, illuminance and light spectrum measurements in a study hall - corridor interior. The supplementary fluorescent electric lights are dimmed down, when the solar fiber system provides the light and the energy savings have been evaluated.

The six fiber luminaires of the solar system provide intensive white light and a high luminous flux. In total 4500 (3000) lm has been obtained under 130 000 lx direct visible sun radiation at a 10 (20) m distance respectively from the light collecting panel of the system. The illuminance level > 500 lx (> 300 lx) has been obtained for a 7 m2 (13 m2) area in the study hall area under the six solar fiber luminaires of the 10 m system during direct visible sun radiation over 100 000 lx. The spectrum of the solar fiber light contains all the visible wavelengths and the color temperature (5800 ± 300 K) is close to that of the direct sunlight as measured outside (ca. 6500 K). High color rendering index (86) has been obtained for the solar fiber luminaires, higher than for the original fluorescent lights (77) in the test area. This high quality light improves the color rendering and fidelity of the illumination and thus improves the visibility of all kinds of objects compared to the fluorescent lights.

The lighting energy saving, as estimated from the test data gathered during the summer, would be 19% in a study hall interior in Uppsala due to the 1790 annual sunny hours, while in southern Europe with 3400 sunny hours it would be at least 46%. The energy used by the solar light system itself and the energy needed to automatically control the electric lighting has been taken into account. In addition, our test data indicates that ventilation energy can be saved during hot summer days.

In corridors and other interiors, where lights are typically on 16 h per day every day, all the sunny hours would occur during the service time, and the savings would be 27% in Uppsala and 55% in Italy. The highest saving of lighting energy, 42% in Uppsala, would be reached in offices assuming that one system illuminates 6 one person offices and the electric lights are switched off during the sunny weather.

It can thus be concluded that the illuminance performance, light quality, energy saving and light balancing with the fluorescent lights have been investigated on a solar fiber optic lighting system at Uppsala in Sweden. A high light quality, good illumination performance and a high efficacy of the lighting system have been obtained. The system can provide the lighting instead of the electric lights, and thus save lighting energy, during sunny weather.

Acknowledgments

D. Johansson, R. Hallqvist, D. Ramstedt and N. Nilsson from Parans Solar Lighting AB, Göteborg are warmly acknowledged for helpful discussions and mutual collaboration. Dr. C. Puglia, The Energy House project of the Uppsala universities; M. Wieselblad, Akademiska Hus AB and L-G Karlsson and M. Magnusson, Bravida, are all very warmly thanked for the financing and installation of the solar light system and all other test site equipment. The European Research Fund is also warmly acknowledged for their support. The EU FP7-NMP project Clear-Up is acknowledged for the funding and support of the test and analysis work. Prof. G. Niklasson is warmly thanked for the useful discussions of this report.

References and links

1.

T. Nakamura, “Optical waveguide system for solar power applications in space,” Proc. SPIE 7423, 74230C, 74230C-10 (2009). [CrossRef]

2.

A. Tsangrassoulis, L. Doulos, M. Santamouris, M. Fontoynont, F. Maamari, M. Wilson, A. Jacobs, J. Solomon, A. Zimmerman, W. Pohl, and G. Mihalakakou, “On the energy efficiency of a prototype hybrid daylighting system,” Sol. Energy 79(1), 56–64 (2005). [CrossRef]

3.

F. Francini, D. Fontani, D. Jafrancesco, L. Mercatelli, and P. Sansoni, “Solar internal lighting using optical collectors and fibres,” Proc. SPIE 6338, 63380O, 63380O-8 (2006). [CrossRef]

4.

Himawari solar fiber optic lighting systems, http://www.himawari-net.co.jp/e_page-index01.html.

5.

T. T. Volotinen, N. Nilsson, D. Johansson, J. Widen, and Ph. Kräuchi, “Solar fibre optic lights -daylight to office desks and corridors,” Proceedings CISBAT2011, 491–496 (2011).

6.

L. C. Maxey, M. V. Lapsa, P. Boudreaux, D. D. Earl, J. Morris, and T. Bunch, “Hybrid Solar lighting: Final technical report and results of the field trial program”, U.S. Department of Energy (DOE) Information Bridge, http://www.osti.gov/bridge, ONRL-TM-150, (2008).

7.

M. S. Mayhoub and D. J. Carter, “Towards hybrid lighting systems: A review,” Lighting Res. Tech. 42(1), 51–71 (2010). [CrossRef]

8.

M. S. Mayhoub and D. J. Carter, “Hybrid Lighting systems: performance and design,” Lighting Res. Tech. 44(3), 261–276 (2012). [CrossRef]

9.

The earlier version (SP2) of Parans Solar lighting system, www.parans.com.

10.

A. Kribus, O. Zik, and J. Karni, “Optical fibers and solar power generation,” Sol. Energy 68(5), 405–416 (2000). [CrossRef]

11.

CIE 84–1989 Photocopy Edition 1996, “The measurement of luminous flux”, Technical report, CIE, Austria (1996).

12.

CIE 15:2004 3rd Edition, “Colorimetry”, Technical report, CIE, Austria (2004).

13.

CIE13, 3–1995, “Method of measuring and specifying colour rendering properties of light sources”, Technical report, CIE, Austria (1995).

14.

C. S. McCamy, “Correlated color temperature as an explicit function of chromaticity coordinates,” Color Res. Appl. 17(2), 142–144 (1992). [CrossRef]

15.

M.-C. Dubois and Å. Blomsterberg, Energy saving potential and strategies for electric lighting in future North European, low energy office buildings: A literature review. Energy and Buildings (2011), doi: [CrossRef] .

16.

Y. Yang and X. Shi, “Analysis and evaluation on luminaire efficacy and colour quality of LED downlights”, Proc. of CIE2012 Lighting quality and energy efficiency-conference CIE x037:2012, pp 05, 546–554, (2012).

17.

http://www.lighting.philips.com/main/led/index.wpdandhttp://www.osram.com/osram_com/products/led-technology/indoor-led-luminaires/index.jsp

18.

L. Edwards and P. Torcellini, A literature review of the effects of natural light on building occupants”, National Renewable Energy Laboratory, Colorado, USA, NREL/TP-550–30769 (2002).

19.

Swedish Meteorological and Hydrological Institute (SMHI), http://www.smhi.se/klimatdata/meteorologi /stralning/normal-solskenstid-for-ett-ar-1.3052, 2011.

OCIS Codes
(060.2310) Fiber optics and optical communications : Fiber optics
(250.5460) Optoelectronics : Polymer waveguides
(350.6050) Other areas of optics : Solar energy
(150.2945) Machine vision : Illumination design

ToC Category:
Illumination Design

History
Original Manuscript: January 24, 2013
Revised Manuscript: March 23, 2013
Manuscript Accepted: April 1, 2013
Published: May 24, 2013

Citation
David Lingfors and Tarja Volotinen, "Illumination performance and energy saving of a solar fiber optic lighting system," Opt. Express 21, A642-A655 (2013)
http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-21-S4-A642


Sort:  Author  |  Year  |  Journal  |  Reset  

References

  1. T. Nakamura, “Optical waveguide system for solar power applications in space,” Proc. SPIE7423, 74230C, 74230C-10 (2009). [CrossRef]
  2. A. Tsangrassoulis, L. Doulos, M. Santamouris, M. Fontoynont, F. Maamari, M. Wilson, A. Jacobs, J. Solomon, A. Zimmerman, W. Pohl, and G. Mihalakakou, “On the energy efficiency of a prototype hybrid daylighting system,” Sol. Energy79(1), 56–64 (2005). [CrossRef]
  3. F. Francini, D. Fontani, D. Jafrancesco, L. Mercatelli, and P. Sansoni, “Solar internal lighting using optical collectors and fibres,” Proc. SPIE6338, 63380O, 63380O-8 (2006). [CrossRef]
  4. Himawari solar fiber optic lighting systems, http://www.himawari-net.co.jp/e_page-index01.html .
  5. T. T. Volotinen, N. Nilsson, D. Johansson, J. Widen, and Ph. Kräuchi, “Solar fibre optic lights -daylight to office desks and corridors,” ProceedingsCISBAT2011, 491–496 (2011).
  6. L. C. Maxey, M. V. Lapsa, P. Boudreaux, D. D. Earl, J. Morris, and T. Bunch, “Hybrid Solar lighting: Final technical report and results of the field trial program”, U.S. Department of Energy (DOE) Information Bridge, http://www.osti.gov/bridge , ONRL-TM-150, (2008).
  7. M. S. Mayhoub and D. J. Carter, “Towards hybrid lighting systems: A review,” Lighting Res. Tech.42(1), 51–71 (2010). [CrossRef]
  8. M. S. Mayhoub and D. J. Carter, “Hybrid Lighting systems: performance and design,” Lighting Res. Tech.44(3), 261–276 (2012). [CrossRef]
  9. The earlier version (SP2) of Parans Solar lighting system, www.parans.com .
  10. A. Kribus, O. Zik, and J. Karni, “Optical fibers and solar power generation,” Sol. Energy68(5), 405–416 (2000). [CrossRef]
  11. CIE 84–1989 Photocopy Edition 1996, “The measurement of luminous flux”, Technical report, CIE, Austria (1996).
  12. CIE 15:2004 3rd Edition, “Colorimetry”, Technical report, CIE, Austria (2004).
  13. CIE13, 3–1995, “Method of measuring and specifying colour rendering properties of light sources”, Technical report, CIE, Austria (1995).
  14. C. S. McCamy, “Correlated color temperature as an explicit function of chromaticity coordinates,” Color Res. Appl.17(2), 142–144 (1992). [CrossRef]
  15. M.-C. Dubois and Å. Blomsterberg, Energy saving potential and strategies for electric lighting in future North European, low energy office buildings: A literature review. Energy and Buildings (2011), doi:. [CrossRef]
  16. Y. Yang and X. Shi, “Analysis and evaluation on luminaire efficacy and colour quality of LED downlights”, Proc. of CIE2012 Lighting quality and energy efficiency-conference CIE x037:2012, pp 05, 546–554, (2012).
  17. http://www.lighting.philips.com/main/led/index.wpd and http://www.osram.com/osram_com/products/led-technology/indoor-led-luminaires/index.jsp
  18. L. Edwards and P. Torcellini, A literature review of the effects of natural light on building occupants”, National Renewable Energy Laboratory, Colorado, USA, NREL/TP-550–30769 (2002).
  19. Swedish Meteorological and Hydrological Institute (SMHI), http://www.smhi.se/klimatdata/meteorologi /stralning/normal-solskenstid-for-ett-ar-1.3052, 2011.

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.


« Previous Article  |  Next Article »

OSA is a member of CrossRef.

CrossCheck Deposited