Optical fiber daylighting systems use large area optical components to collect direct sunlight and transfer the light to interiors through optical fiber. This technology has the ability to bring sunlight much deeper into buildings, which is a major obstacle for most daylighting strategies to overcome before they can reduce the dependence on artificial lighting. As a type of green energy, daylighting is essential for improving environments and reducing electric lighting power consumption in office buildings [1]. Efficient daylighting buildings are estimated to reduce electric lighting energy consumption by 50-80%. Optical fiber daylighting also decreases heating, ventilation, and air conditioning (HVAC) costs because it generates no heat. In terms of health, artificial light fails to produce a comfortable indoor environment. Daylight can be used to reduce the impact of illnesses such as seasonal affective disorder, to improve worker productivity [2]. Research clearly shows that daylighting improves the quality of life and saves electric energy usage. However, most existing optical fiber daylighting systems do not provide a cost-effective solution, therefore it is difficult for the systems to be widely used and to become successful in the market. In conventional optical fiber daylighting systems, a point focusing Fresnel lens or parabolic mirror is generally used. The incident sunlight falling on a large lens/mirrors area is focused on one end of a single or bundle of optical fibers. However, this design suffers from the non-uniformity of the light beam over the end of optical fibers, therefore an additional secondary optics is needed to homogenize sunlight and increase the tracking tolerance [3]. Different techniques have been proposed in the literature to capture sunlight through reflectors and lenses. Traditionally, optical fiber daylighting systems were implemented only on a small scale. For scaling up a daylighting system, a large number of concentrators were used to focus the light into optical fiber bundles. However, these systems become costly when the number of sunlight capturing modules increases [1]. When the daylight is inadequate or not available, a light emitting diode (LED) lighting system is used to compensate daylight. The LEDs can be turned on and dimmed through closed-loop control when the daylight illuminance is inadequate. The daylight part and LEDs are combined within a specific luminaire but it is very difficult to provide stable and uniform illumination because they are two independent lighting systems. Combining two independent lighting systems also creates more cost.

In fact, optical fiber daylighting is already a commercial technology [4-6] and different prototypes of this method have been proposed [7-11]. Most existing optical fiber day-lighting systems use visible range only - a small fraction of the sunlight spectrum - for illumination. The non-visible range is not suitable for lighting and it produces heat. An additional module is required to remove heat. But if a concentrating photovoltaic system (CPV) is integrated in the system, the un-useful non-visible light can be directly converted to electric power that can be utilized as the LED power supply, and the complicated heat removing module can be eliminated. Concentrating photovoltaic (CPV) technology is an effective means towards improving the energy supply and reducing the use of solar cell material. A CPV system utilizes the optical concentrator to focus sunlight to the small size of solar cell. As a conventional concept, Chih-Hsuan Tsuei et al. have investigated an optical fiber daylighting system integrated with CPV [12]. This system consists of two reflectors as sunlight concentrator, one beam splitter, an optical switching system for saving solar energy, and a light guide. The schematic diagram of an original optical fiber daylighting system combined with CPV and LED is shown in Fig. 1(a). However, it is difficult to scale up this kind of system because they utilized the reflectors as primary sunlight concentrator as discussed above. Besides that, daylighting module and LED illumination module are separated, so that they do not provide stable and uniform illumination in the interior. Two independent modules also make the systems costly. There is no space to integrate LEDs inside the daylighting module to make a compact system.

In this study, we demonstrated a design for large scale hybrid optical fiber daylighting/LED system using Fresnel lenses and stepped thickness waveguides (STWs) that can utilize the full spectrum range of sunlight. The original concept of utilizing a kind of planar solar concentrator of “stepped-shape” to collect and redirect sunlight from multi Fresnel lenses was proposed in [13]. The system in [13] used linear Fresnel lenses and a larger slab STW. Sunlight was focused from a large area to a line by linear Fresnel lenses so that the system can use only a single sun-tracking system instead of the usual dual sun tracking system. Although the single sun-tracking system can reduce cost, line focusing from linear Fresnel lens does not provide high concentration ratio. Low concentration ratio will require a large number of optical fibers in light guiding module and will make the system costly. On the other hand, line focusing also does not support a solution to integrate a CPV module inside the sunlight concentrator. This is the reason why we introduce an alternative approach by using Fresnel lens and narrow STW in this research. Point focusing from Fresnel lenses can increase the concentration ratio of the system and support a good idea to integrate CPV module and LED illumination inside the sunlight concentrator. This approach is able to make the system more compact. The schematic diagram of the system is shown in Fig. 1(b). The sunlight collector of the system is a one dimensional Fresnel lenses array. Light collected by each element of the Fresnel lenses array passes through a plate of beam splitters (PBS) to be separated into visible and non-visible rays [12]. Visible rays are coupled into an STW and propagate within the waveguide by total internal reflection (TIR). And finally they are coupled to a plastic optical fibers (POFs) bundle placed at the exit port of the waveguide [14]. The non-visible rays are absorbed by the solar cells to produce electricity which can be stored on solar energy storage (SES) like a battery. The SES is used as power supply for high power LEDs that are integrated at the other end of the stepped thickness waveguide. The LEDs’ light is also guided in the waveguide by TIR and combined with the visible sunlight in the POF’s bundle for uniform illumination in the interior through distributing lenses. An optical sensor and lighting control system are used to control the LED light flow to ensure that the total output flux is a fixed value when the sunlight is inadequate.

The remainder of the paper is organized in the following manner. In Section 2, designs for utilization of sunlight concentrators, light guides and solar cells are discussed. In Section 3, the optical fiber daylighting system based on STW was modeled in LightTools^TM to evaluate the performance of the system. Finally, brief concluding remarks and future work are included in Section 4.

A hybrid optical fiber solar/LEDs lighting system powered by renewable solar energy for indoor illumination is presented in this section. First we designed a sunlight concentrator to collect direct sunlight [12]. The sunlight concentrator, as shown in Fig. 2, is composed of a linear array of five Fresnel lenses, five beam splitters, STW and POFs bundle. The concentrator is supposed to be equipped with a tracking system to collect sunlight in the normal direction. The visible light passes though the PBS and is coupled into an STW using specular reflections from an associate area of mirror facets fabricated at each lens focus. The facet redirects light into a guide layer. Rays that exceed the critical angle defined by Snell’s Law propagate via total internal reflection (TIR) within the waveguide to the exit aperture. STWs also provide excellent beam homogenization when coupling diverging illumination into a high number of supported modes [15]. The waveguide transports sunlight collected over the entire input aperture to a bundle of POFs placed at the exit port of the waveguide [16]. While the non-visible rays reflected at the surface of PBS are collected by solar cells. The components of the optical system, design parameters and their effects on the optical performance are discussed below in detail.

The primary concentrator array can be constructed from several different types of point focus concentrators such as fly eye lens array, array of Fresnel lenses, Compound Parabolic Concentrators (CPCs), diffractive concentrators or combinations of lenses and mirrors. Because of the expensive cost and quality degradation, conventional lenses and high reflectivity mirrors are not recommended for the primary concentrator of the daylighting system. In this study, the primary concentrator is constructed from one dimensional Fresnel lenses array. A non-imaging Fresnel lens uses ring-shaped segments and its cross sections are straight lines rather than circular arcs. Such a lens can focus light on a small spot, but does not produce a sharp image. These lenses have application in solar power, such as focusing sunlight on a solar panel. We employed low cost commercial Fresnel lenses made by DiYPRO Co., Ltd. (Korea) with the performance data as in Table 1 [17].

In our optical design and simulations, five Fresnel lenses with the size of 240 mm × 240 mm are arranged linearly as shown in Fig. 2 so that the total length of the system is L = 1200 mm (Fig. 3). The focused regions are five points which are also arranged in the linear array on the focal plane. However, the Fresnel lenses also have several inherent disadvantages. One of them is the dispersion of the solar spectrum. For daylighting systems, the dispersion of visible range is very important since it leads to an essential decrease of the optical efficiency and concentration ratio of the systems, especially it is important for the Fresnel lens with high concentration ratio. The concentration ratio of primary lenses is calculated by ratio of focused area and lens area. We used ray tracing in LightTools^TM to analyze the dispersion of visible range. The sunlight source used in the analysis is in the range of 400 - 750 nm. Figure 3 shows the sunlight distribution on the focal plane. Figure 3 also illustrates schematically how the sunlight beam passes through a Fresnel lens. It is noted that the dimensions of some specialty parts of the system are locally exaggerated for convenient explanation. The sunlight is dispersed at the focal plane of the Fresnel lens due to the wavelength dependence of the refractive index of the lens material. We found that the light was focused from each Fresnel lens to a small spot with diameter d = 8 mm. The focused area also defines how big the sun image will be at the focal region, and it is a design parameter for directing surfaces and light guide thickness, which will be discussed in detail in the next Section 2.2.

The focal length of each Fresnel lens in the primary concentrator array defines the main thickness of the total structure. In this case, the focal length was 300 mm (Table 1), which means the thickness H of the concentration part is around 300 mm. Together with focal length, the aperture of each Fresnel lens is an important parameter. The lens aperture is usually specified as an f-number, the ratio of focal length to effective aperture diameter. For the Fresnel lenses in this study, the f-number is f/l = 1.25. If the f number gets smaller, then some light rays coming to the focus are refracted too much. The refaction angle is an important parameter for the directing surface and should be properly designed to achieve total internal reflection at the directing surface and the walls of the light guide structure [14]. The focused area and the f-number are most important input parameters for STW that will be discussed below.

This waveguide is responsible for redirecting and transferring the light to the exit port of the concentrator. As shown in Fig. 1, this part stays below the primary concentrating lens array. Light directing surfaces, light transmitting media and the exit port of the concentrator are actually features of the horizontal STW structure (Fig. 4). Completely efficient waveguide coupling from multiple locations and lossless propagation can only occur through a monotonic increase in modal volume such as STW [15, 17].

The length of the waveguide is determined by the distance between the last focal point and the exit port, i.e. l = 1080 mm as shown in Fig. 4(a). Light directing surfaces are mirror coating facets that reflect the light into the waveguide with an angle such that reflection from waveguide surface satisfies the TIR condition. From f-number = 1.25, we calculated and found out that the directing mirrors are 30o inclined surfaces, as show in Fig. 4(b). Since the diameter of light focused area of each lens is d = 8 mm so the size of the directing mirror must be larger than 8 mm to cover all sunlight input. But increasing the width of the directing mirror will increase the thickness and the width of waveguide and reduce the concentration ratio of the system. In this situation, appropriate area of directing mirror is a square 9 mm × 9 mm and the thickness of one step of waveguide is s = 4.5 mm. So the end of stepped-thickness waveguide with five steps (exit port height) is h = 22.5 mm. With f-number =1.25, the light that strikes the side surfaces of STW also satisfies the TIR condition. The is no leak of light at the side surfaces. As shown in two inserts of Fig. 4(b), the sunlight which is reflected by light directing surface is transferred to the exit port by TIR at the front surface, bottom surface and two side surfaces of the STW. Because the sunlight is confined inside the STW by TIR, so the reflection loss can be ignored. Only the absorption of the stepped thickness waveguide is a major loss inside the STW and it can significantly affect the final efficiency of the system. The total losses of the system will be discussed in the next section. In accordance with the arrangement of the proposed solar concentrator, as shown in Fig. 2, an important property of the stepped thickness solar concentrator is a geometric concentration ratio Cgeo given by:

where, D and L are width and length of Fresnel lens array as shown in Fig. 3; w and h are sizes of exit port as shown in Fig. 4(a). In this case the concentration ratio Cgeo is 1422.

The material absorption is also important and it can significantly affect the final efficiency of the system because the 1080 mm length of the proposed STW is quite long. As the light travels considerably longer distance inside the light guide, the absorption by the light guide material becomes important. Therefore, a low absorbing material should be used for the STW.

Visible light in its most fundamental definition is electro-magnetic waves whose wavelengths range from 380 nm to 780 nm. The basic source of light on earth is the sun which also emits electromagnetic waves in the ultra violet (UV), visible light, near infrared (NIR) and long IR wavelength (heat) part of the electromagnetic spectrum. In every optical fiber daylighting system, if daylight is to be transmitted over plastic optical fiber, no-heating is desirable. In the traditional daylighting systems, cooling is a trouble making issue at high concentration levels. Sophisticated active and passive coolers are needed to remove the heat generated by the focused sun light at high concentrations [14]. In most of the cases, the plates of the beam splitter are used to remove heat. PBS is a kind of filter which reflects the ultraviolet (UV), and near infrared (NIR) parts of the solar light. In our proposed design, PBS is a spectrally selective hot mirror from Edmund Optics Inc. [19]. The hot mirror features a multi-layer dielectric coating optimized for greater than 98% transmission of visible light and greater than 95% reflection of IR wavelengths. The variation of transmission and reflection of the plate beam splitter according to wavelength is shown in Fig. 5.

The PBSs are placed above the focus of the Fresnel lenses. The infrared rays are reflected by PBSs and are utilized by photovoltaic cells to generate electricity (Fig. 5). The solar cells used in this study must be specially sensitive to UV and NIR wavelengths. Considering the spectrum of absorption and illumination at the same time, CuInSe (CIS) thin film photovoltaic devices are the best candidate for this purpose [12]. CIS solar cell (or photovoltaic) is basically a p-n diode. This structure can convert sunlight directly into electricity with a maximum reported laboratory scale cell efficiency of ~20%. The CIS, a direct band gap material, can absorb sunlight efficiently from 280 nm to 1200 nm. A collimating lens is used to ensure uniformity of the concentrated sunlight to avoid damage to the solar cell. The visible rays (400 nm to 700 nm) are designed for indoor lighting; the non-visible rays (280 nm to 400 nm and 700 nm to 1300 nm) for solar power generation [12].

The optical fiber consists of core, cladding, and external protective coating. The light travels inside the core, and the cladding, which has a lower refractive index, provides internal reflection at the boundary of the core. The optical fibers used in the daylighting and solar thermal applications for the transmission of sunlight should transmit a broad spectrum. One of the most significant features of sunlight transportation is the wiring method, and the wiring must be as simple as electrical wiring. Therefore, only optical fibers are capable of fulfilling the requirement. We used optical fibers to deliver sunlight to the interior with a small amount of loss. Silica optical fibers (SOFs) are known as good light-transmission media and have the best resistance to heating; however, they are expensive. Plastic optical fibers (POFs) have substantially higher attenuation coefficients than SOFs but POFs are preferred in daylighting systems due to their low cost of production, tighter minimum bend radii, ease of installation, and durability for complex wiring in buildings. In this design, we employ the large core POFs with 2 mm core diameter from Edmund Optics [20]. The light can be transferred over long distance without visible changing of the input color because the POFs are made of PMMA. They have attenuation minima of 64, 73, and 130-dB/km occurring at 520, 570, and 650 nm, respectively. These wavelengths indicate that PMMA fibers will transmit green, yellow and red light particularly well. The fiber parameters are listed in Table 2.

Besides attenuation of light within the optical fiber, coupling losses account for the most significant reductions in transmitted light intensity. The size of exit port from concentrator is 9 mm × 22.5 mm (Fig. 4(a)) and the optical fiber diameter is 2 mm, so a ribbon configuration of POFs is proposed for optical fiber coupling as shown in Fig. 7(a). In this design, forty four POFs are required and arranged in an 11×4 rectangular array. The geometric coupling efficiency is calculated by ratio of cross section area of optical fiber ribbon and exit port area, ηgeo-coupling = 0.75. Multiple optical fibers were connected to the waveguide by polishing the ends of bundle fibers and simply holding the polished ends to the waveguide as tightly as possible. Intensity reductions occur primarily as a result of imperfections in end-face geometry and numerical aperture mismatch as well as Fresnel reflection losses. Fresnel reflection losses occur when light exits the waveguide material and enters an optical fiber medium with a different refractive index, such as air. Fresnel losses are relatively constant no matter what distance is involved, resulting in an approximately four percent loss of light intensity leaving the waveguide and an additional four percent reduction in remaining light intensity upon entering the POFs bundle, even when the air gap is microscopically small. To reduce Fresnel losses due to air gap between waveguide and endface of POFs, index matching gel was applied to fill the air gap. The numerical aperture mismatching between optical fiber and stepped index waveguide also causes significant losses. The optical fiber and the stepped thickness waveguide in this case have a maximum acceptance angle θ, given in the following relationship:

where, n_co and n_cl are the refractive indices of the core and cladding, respectively; n_o is the refractive index of the matching gel. By taking the values in Table 2, we can find that the acceptance angle of plastic optical fiber is θ = ±19°. The input angle of light that came from exit port of the waveguide to optical fibers can be calculated based on the numerical aperture of the waveguide. Using Eq. (2) with low index of 1 (air) and high index of 1.49 (index of PMMA), the angle of sunlight rays escape from the waveguide exit port sweep in the range of ±49° (output angle of rays from waveguide also illustrated in Fig. 4(b) by ray tracing method using LightTools^TM software), so only some rays of the output beam which have angle less than ±19°, can couple to optical fibers. To solve this challenge, an additional inverted dielectric extruded compound parabolic concentrator (CPC) can be attached to the end of the waveguide slab as shown in Fig. 7(b), causing the light to collimate in a larger area. CPC is a non-imaging optical element that was originally developed for efficiently concentrating collimated light into a smaller area. Conversely, it can also be used as a collimator by using its minor aperture as input. The collimated CPC should be designed to achieve an appropriate degree of collimation to facilitate an effective fiber-coupling performance. The diameter of the conjunction area between the waveguide exit port and collimated CPCs is the same 9×22.5 mm. The minor aperture angle needs to be fixed to the 49° to capture all of the output light from the waveguide exit port. The primary aperture angle should be 19°, the same numerical aperture of optical fiber. 3D structure of extruded CPC is generated by LightTools^TM software with the above input structure parameters as shown in Fig. 7(b). The primary aperture of CPC size which corresponds with exit port is 15 mm × 30 mm. The total number of optical fibers required to fill the exit port area is increased to 105 and fibers are arranged in a 15×7 rectangular array.

By adding CPC as collimated device for optimal fiber coupling, the concentration ratio of the system becomes decreased from 1422 to 640 (concentration ratio, Cgeo is determined by Eq. (1) with replacing waveguide exit port area by primary aperture of CPC). The influence of collimated CPC to the optical efficiency of the system and evaluation of the advantage of CPC will be discussed in Section 3.

Conventional hybrid optical fiber daylighting/LED systems are comprised of two independent modules: daylighting module and LED lighting module. Double set of luminaires has to be used. The method of providing the uniform light distribution for LED light and sunlight at the same time has not been viable. In this section, we propose an integrated hybrid optical fiber daylighting/LED system in which LEDs are coupled to a stepped thickness waveguide such that light is confined by TIR defined by Snell’s law. Our proposed systems can provide uniform and stable light distribution inside the building all the time even when the amount of daylight fluctuates. The supplementary LED light source controls its intensity according to the amount of daylight change. White LEDs currently have superior luminance and efficacy compared to other broadband sources. Consequently, LEDs with the highest luminance are desirable because they provide more optical power with the same étendue. High-performance LEDs typically have a lens positioned right on top of the emitting surface. With this lens, it is possible to emit more light from the LED. Unfortunately, this lens also prohibits attaching an optical component directly to the LED. Typical LEDs have a very broad solid angle of emission whereas waveguides typically have only a very small acceptance angle for light to be coupled into. Therefore the coupling efficiency between an LED and a waveguide decreases as the distance between the components increases [21]. However, LED Chip-on-Board (COB) technology has developed quickly recently and it gives us a solution to solve this problem. In this study, we optimized certain aspects of the design space using parameters from commercially available LEDs-COB of the ProPhotonix company. The COB LED technology describes the mounting of a bare LED chip in direct contact with the substrate to produce LED arrays. It is a method of LED packaging which has a number of advantages over traditional surface mount technologies. Due to the small size of the LED chip, COB technology allows much higher packing density than surface mount technology. This results in higher intensity and easy coupling with the waveguide. We employed a one dimensional array of five LEDs on board to couple with the waveguide. LED-STW coupling configuration is shown in Fig. 8(a). A small piece at the end of STW was removed to make a surface for LED coupling. The end of the waveguide came to have the shape of a truncated cone. The emitting surface of the LED array was positioned in the coupling surface and fixed by optical epoxy. The light emitted from LED was confined inside the waveguide by TIR law as shown in Fig. 8(b). The optical performance of the LEDs-waveguide was analyzed with Monte Carlo ray tracing using LightTools^TM. The LED-STW coupling efficiency is simply defined as the ratio of the output luminous flux at the exit port of STW to the luminous flux which emitted from LEDs. With the design described here, the coupling efficiency was found to be 98%. To the best of our knowledge, the hybrid optical fiber daylighting/LED systems that is described here is the first system with LED integrated inside the sunlight concentration part. This proposed design allows LED and daylight to use the same infrastructure such as optical conduits and lighting distribution systems in the interior, which reduces system cost significantly.

Optical modelling plays a crucial role in the efficiency evaluation of an optical system. Commercial optical modeling software, LightTools^TM, was used to design and simulate the geometrical structure of the stepped-thickness waveguide optical fiber daylighting system. In the designed system, one of the most common optical plastic, poly-methyl methacrylate (PMMA) with refractive index of nPMMA = 1.49 is selected for Fresnel lenses, stepped index waveguide and optical fiber. The Fresnel lenses were commercial products from DiYPRO Co., Ltd. (Korea) with parameters in Table 1 and the POFs were Anchoroptics Ltd. products with parameters in Table 2. The coupling region between the exit port of the waveguide and the optical fibers was filled by epoxy as matching gel with refractive index of n=1.51. To evaluate the losses in the system, in simulation model, we inserted two luminous flux receivers as shown in Fig. 9(a). The simulation for optical efficiency of the waveguide concentrator was implemented for both cases: with and without collimated CPC attached at the exit port, to evaluate the advantage of CPC. The optical efficiency is simply defined as the ratio of the output luminous flux to the input luminous flux. With the system parameters given above, the system efficiencies were achieved to be 35% in the case without collimated CPC and 63.8% in the case including CPC. With the collimated CPC, the optical efficiency was improved significantly because the angle of exit light satisfied the numerical aperture condition of POFs. The increasing of optical efficiency lowers the geometric concentration ratio from 1422 to 640. However, comparing with available commercial systems, the large scale system with concentration ratio 640 is deemed reasonable. The uniform distribution of light flux at the output is an important factor of daylighting system. In this proposed solar concentrator, the STW also serves as homogenizer for attaining good irradiance uniformity in the output aperture. Inside an STW, most of the sunlight reaching the optical fiber after bouncing off the waveguide surface by TIR and the light distribution on output aperture can be made uniform. As is evident in Fig. 10, uniform illumination was achieved at the endface of POFs bundle.

Since the ultra violet (UV), near infrared (NIR) and long IR wavelength (heat) part of the sunlight spectrum were filtered by PBS, the heat will be dissipated and does not influence the stepped-thickness waveguide and POFs.

The optical efficiency of LED coupling with STW, the other important part of our proposed system, is also simulated and calculated in this section. LEDs are used to maintain the illumination levels when the daylight falls from the required value. An electric lighting control system is used to maintain LED light. The basic principle of the hybrid system is that a photocell senses available light at the output of the concentration system, and a control algorithm maintains an average luminous flux value at the output of concentrator. A schematic diagram of this process was introduced in Fig. 1(b). In our proposed concept, LED was integrated at one end of the STW as shown in Section 2.5. We used OSRAM LUW-W5AM LED chip, which had a electric/optical convection efficiency of 130 lm/W. Simulation configuration is described in Fig. 9(b). The optical efficiency of LED coupling is calculated by the ratio between total luminous flux measured at all optical fiber endfaces and luminous flux emitted from LEDs. With the simulation parameters given above, LED coupling efficiency is found to be 74.5%, including Fresnel loss, geometrical mismatch and optical fiber coupling losses.

Proper operation of the proposed hybrid optical fiber daylighting system requires direct sunlight focused into the center of directing mirror facets, which is a difficult task because of the sun’s position changing in the sky. Tracking systems have been developed with the goals of maintaining precise alignment with the sun, maximizing operational time, and thus minimizing the power required for electric lighting. Maintaining proper collector alignment with the sun requires a mechanical tracking system capable of tracking both vertically and horizontally relative to the earth [22]. The required accuracy of the sun tracking system is determined by a solar concentrating collector’s angular tolerance. The angular tolerance of the system is the acceptable angular deviation of the sunlight direction from the two main axes of the system within the allowable efficiency loss. It is defined as the angular deviation where efficiency drops 90% of its maximum value. We examined the optical efficiency with a different angular deviation of sunlight direction along the North-South (NS) and East-West (EW) directions. The alignment of system along NS and EW directions is indicated in Fig. 11. Figure 12 shows that the optical efficiency depends on angular deviation along NS and EW direction. In the simulated system we studied, the results show that the tolerance is more than ±0.75° along NS-direction and ±0.50 along EW-direction. The usual tolerance for conventional systems is less than ±0.2° [21]. Therefore, the proposed system can lower the accuracy requirement for a sun tracking system in comparison with other available commercial daylighting systems, which eventually reduces cost.

As each thickness step or width of waveguide gets larger, the focused light from the primary lens has more space to move without decoupling from the light guide media. Therefore, enlarging volumetric increase of light guide while keeping the primary concentrator size fixed, gives better tracking tolerance. However, increasing the acceptance angle reduces the concentration ratio dramatically. In system design, the tolerance and the concentration issues should keep in balance.

As an initiative for a cost-effective large scale optical fiber daylighting system, five stepped thickness daylighting systems are placed together. These systems share sun tracking system, battery bank and other infrastructure. The system will include a 5×5 array of Fresnel lenses and a linear array of five STWs. A schematic depiction of a sunlight collector for a large scale fiber optic daylighting system is shown in Fig. 13.

To evaluate the performance of the large scale daylighting system, the illuminance from the sunlight was measured at different times. The site of application was located at 127.0 longitude and 37.50 latitude. Here let us look at the illuminance for a summer day as an example. The highest solar elevation (zenith) angle at the site is 760, and it takes place at 12:30 PM [1]. To achieve direct sunlight, we assume that the daylighting system has a sun tracking device to rotate the concentrator module toward the sun all times of the day. The area of the sunlight collector is 1.44 m2 because this is the array of 5×5 Fresnel lenses with size of 0.24 m × 0.24 m. The conversion into luminous flux in lumens can be calculated by,

where E is the measured illuminance of sunlight, S is the area of sunlight concentrator, and F is the input luminous flux in lumens. With Eq. (3), we can calculate the simulated input flux from the measured illuminance outdoors to calculate total output at the end of POFs, and this is the indoor illumination [12]. The measured illuminances, input luminous fluxes at the surface of the concentrator are listed in Table 3. Daylighting systems with optical fibers are being used to guide light deep into a building, therefore, the transmission loss of POF should be included in the calculation. If we assume that the daylight capturing system is placed on the top of the building to illuminate a room at 6 m below from the top, roughly 10 m of POF is needed from the capturing point to the distribution area. The transmission loss of 2 mm core diameter POF is 0.46 dB/m. This means that the luminous flux decreases by 2.8 times when propagating along 10m length of POF. The loss because of attennuation in PBS is estimated by 2% based on dependence of transmittance on the wavelength in Fig. 5. To achieve uniform illumination from optical fiber, a diverging lens was the best choice for light distribution [23]. The loss at the light distribution module in the interior is only the Fresnel loss on the surface of distributing lens. It is about 4% for air/glass interface. The angular acceptance along EW direction was ±0.5° so that we can use a typical sun tracking system whose tolerance is lower than ±0.2° [24]. With tolerance of ±0.2°, the loss due to sun tracking system is about 5%. The illumination in the interior (lm) at different times of the day is shown in the last column of Table 3. Losses due to Fresnel reflection, optical fiber coupling, optical fiber absorption, sun tracking, PBS absorption, light distributors were included in this calculation. An office building is required to achieve an average illuminance of 500 lux. At the highest outdoor illuminance at 12 PM, the output flux is about 31000 lm as shown in Table 3. The proposed system can illuminate for 62 m² . In other times of the day, the sunlight in the interior becomes less than the required value of 500 lux, and the LED sources should be turned on to compensate. The variation of luminous flux in the interior at different hours for daylighting, LED lighting and hybrid lighting are calculated and shown in Fig. 13.

The total luminous flux provided by solar lighting and LED lighting during 12 hours of working day can be calculated by Eq. (4).

Here, T is total luminous flux in one day; F is luminous flux of solar lighting or LED lighting at certain time; t is time of the day. Based on Eq. (4) and Fig. 13, the solar lighting can provide 204,321 lm per day and LED should compensate 167,679 lm per day. It is clear that, in a sunny day, solar lighting can provide 54.9% in total light consumption during 12 hours of a working day. A part of electric consumption for LEDs is utilized from CPV module as an additional energy saving efficacy.

Providing a cost-effective lighting system is the major problem that faces any daylighting system. We can evaluate our proposed system in terms of daylight collection capability and cost effectiveness by comparing the system with two other commercialized systems: Himawari and Parans systems. Due to variety of operating conditions, system comparison is not easy. The commercial daylighting system price lists are mostly available only for customers and not for academic purposes, so that details on raw material cost, installation cost and so on were not published. However, some common criteria should be indicated for the comparison purpose such as: systems in comparison must work under same sky condition (sunlight illuminance); same optical fiber length which can determine how deep daylight can be delivered inside a building. We used the data from developers’ publication to estimate our proposed system in accordance with theirs. Since Himawari and Parans systems are daylighting systems without CPV and LED integration, we estimated only the cost of the daylighting module in our proposed system for comparison purposes. To estimate the cost of optical fiber daylighting module of the proposed system, we make the following considerations: the cost of 5×5 array of Fresnel lenses is about $1,000 ($40/lens); 5 STWs’ cost are about $250 ($50/STW). The cost of five POF bundles with 10 m length is about $2,625 (the POF price is $0.5/m, 105 fibers in a bundle). The 25 PBSs have total price of $875 (25 mm square hot mirror from Edmund Optics is $35/sheet). Automatic dual axes sun tracking system has cost of $700, roughly. Total expected cost of the proposed system is $5450. The total luminous flux of a proposed system is 29000 lm which was measured under direct sun illuminance of about 100,000 lux. In average, 1 Klm has a price of $187. The estimated cost for daylighting module is listed in Table 4. The Himawari system collects and concentrates sunlight using multiple sun tracking Fresnel lenses. Light is transported by quartz optical fibers. The total luminous flux of a Himawari system is 4000 lm which was measured under direct sun illuminance of about 100,000 lux. The net price of a Himawari system package (including 12 lens collector, 10 m optical fiber bundle) is $6,240 [25]. The average price of 1 Klm of Himawari systems is about $1605. Swedish Parans system consists of an array of many Fresnel lenses that are mounted on the common sun tracking system and concentrates daylight. Parans system’s total output flux is about 3540 lm, which was measured under direct sun illuminance of 100,000 lux. Parans systems package with 10 m of optical fiber bundle, net price is about $5425 [26]. In average, 100 Klm of Parans system have cost of $1533. The comparisons associated with the daylight collection capability and cost have been summarized in Table 5. As shown in the table, although the price of these systems are almost the same but our proposed system can deliver much more sunlight into the interior, thus average price per 1 Klm of our proposed system is 9 times cheaper than some commercially available systems.

The design concept introduced here fulfills the requirements for a concentrator of a large scale hybrid optical fiber daylighting system, such as a high concentration ratio and uniform distribution of output light. The performance of a large scale of daylighting system realized by integration of five systems was evaluated on a sunny day. Recently, a recognized effective design strategy leads to the simplification of product/system design and reduction of the number of stages in the manufacturing process. Reducing the use of materials will also contribute to reduced manufacturing costs, supplier management costs, and recycling costs. The daylighting system concept proposed here meets all of these requirements. The waveguide, Fresnel lenses, and POFs are made of a single material - PMMA. From the viewpoint of fabrication, STW has very simple structure so that it would be fabricated by an injection molding process [24, 25]. All these features result in a huge reduction of manufacturing cost.

Although most of loss mechanisms of the proposed solar lighting system were included in the simulation, when the system is implemented in a real experiment, some other losses and degradation which cannot be calculated by simulation may appear. Firstly, the efficiency of a solar lighting system strongly depends on the characteristics and performance of the optical components. Therefore, the characterization of the optical components and their imperfectness during the fabrication process becomes an important issue when the proposed design is realized in a practical experiment. Because of imperfection of optical daylighting elements, loss of optical efficiency directly translates into reduced system efficiency. The lenses based on refractive designs such as Fresnel lenses currently dominate the CPV and daylighting industry for modules using substantial geometric concentration of sunlight. Fresnel lenses composed of PMMA are historically low cost implementation means. A Fresnel lens consists of discrete concentric prism elements composed of silicone patterned in a PMMA substrate. The causes of imperfection and degradation of Fresnel lenses are surface crazing, UV, temperature, humidity, and abrasion. For STWs, the imperfection in geometry structure, impurity in the material, and reflectivity of redirecting mirror reduce the optical efficiency of the system. In our proposed system, glass-material components such as lenses, STW, and optical fibers were replaced by plastic materials so that the cost of the system is reduced a lot. However, the degradation due to ultraviolet (UV) radiation and temperature is also a problem in plastic material. The degradation due to UV radiation can be eliminated by using glass optical fibers, however, the cost of glass fibers is five times more than POFs. Fortunately, PBS was used underneath the Fresnel lens array to change direction of UV and IR radiation, which can protect STWs and POFs from UV and heat. Shorter lifetime of POFs also is a challenge for every optical daylighting system. In addition, the alignment of optical elements, optical fiber coupling, integration of LEDs into STW, mechanical design are also technical problems when the system is implemented. In the next study, we will try to implement experimentation under real conditions to verify the accuracy of the simulation, so that we can evaluate the degradation, study reliability concern and the commercial viability of the system.

A hybrid optical fiber daylighting/LEDs system using TIR-based STW as a coupler to optimize the optical efficiency and geometric concentration ratio has been designed and discussed. It was verified that the solar cells module, LEDs lighting, and the optical fiber daylighting are compatible together in a compact system. The full spectrum of sunlight collected by the concentrator is split into visible and non-visible rays by a beam splitter. The visible rays passing through the light guides enter an interior space directly. The non-visible rays are absorbed by the special CIS solar cells to provide electrical power to LEDs. The heat problem is also solved because the heat part of the sunlight spectrum (non-visible rays) is separated away from the optical concentration region. To explore the practical performance of the system, an appropriate design of optical system was proposed, modeled, and simulated with ray-tracing based LightTools^TM software. Simulation results indicate that concentration ratio of C_geo = 630 and optical efficiency of 63.8% were achieved. Moreover, the tolerance of the system along NS and EW axes was also analyzed. The simulation results show that the tolerances were ± 0.75o along NS-direction and ± 0.5o along EW-direction. LEDs lighting module was integrated in the STW instead of standing independently as in the conventional systems. This is a solution to enhance the optical output flux quality and lower the cost. A large scale hybrid daylighting/LED system was realized by integrating five STW systems in a linear array. Most optical elements used in this design such as Fresnel lenses, POFs, LEDs, PBS are commercial products to solve the cost problem. It shows great potential for commercial and industrial scale of daylighting system fields [27].