Lighting Systems Energy Efficiency based on Different Human Visual Conditions

Transcription

1 Lighting Systems Energy Efficiency based on Different Human Visual Conditions J. Fraytag, M. F. da Silva, N. B. Chagas, R. N. do Prado, IEEE Member and M. A. Dalla Costa, IEEE Member Federal University of Santa Maria UFSM Group of Intelligence in Lighting GEDRE Santa Maria, RS , Brazil Abstract- This paper presents a lighting design analysis in scotopic and photopic conditions of different lamps types. The efficient lighting system concept is usually associated with high performance electronic circuits. However, researches related to the human eye physiology indicate that this efficiency gain can be achieved only by changing the lamp type to be used in each visual condition. Analysis suggests that changing the light spectrum incident on the eye, so that it excites most effectively the photoreceptor cells, light levels can be reduced without compromising the visual performance. On this basis, this paper presents a study comparing the power required, the lighting system investment cost, and the energy consumed cost over 10 years. Therefore, lighting systems based on high pressure sodium lamps, fluorescent lamps, metal halide lamp, and lighting emitting diodes are compared. The final results show that the most efficient lighting system for photopic conditions is based on high pressure sodium lamps. However, under scotopic conditions the best efficiency is obtained with the use of metal halide lamps. I. INTRODUCTION The artificial light sources have an important role in creating the necessary conditions for a human being good sight. However, lighting design, as well as lighting efficiency plans, has been drawn only from the perspective of lighting indices, energy efficiency and luminous efficiency (lumen/watt). The use of a suitable lamp for a particular environment can improve the visual efficiency, which is related to color perception, visual acuity, and the speed and accuracy needed to one perform his visual tasks [1]. Research suggests that changing the light spectrum, light levels can be reduced without compromising the visual performance [2]. According to this relationship, the lamps power rate can be significantly reduced, for the same lighting result, increasing the system efficiency without needing to develop complex electrical circuits for this purpose. Regarding economy issues, this represents a considerable reduction in electric energy consumption, which potentiates a higher lighting efficiency. In [3] an analysis of energy efficiency considering HPS (High Pressure Sodium) lamps and MH (Metal Halide) lamps for public lighting is presented. There are several papers in the literature that propose a luminous efficiency analysis considering the human eye sensitivity spectrum and the lamp radiation spectrum. In [4], a comparative study between HPS lamps and LED (Light Emmiting Diode) evaluating the human eye structure and the light radiation influence on it is carried out. Reference [5] conducted a public lighting efficiency review, a comparison between HPS lamps and fluorescent lamps in low light This paper presents a comparative study of HPS lamps, fluorescent lamps, MH lamps, and LEDs light bulbs in scotopic and photopic conditions, considering the human eye sensitivity, in order to determine the most efficient source to be used in each visual condition. Moreover, a case study that examines the relationship between power consumption and initial investment cost over 10 years for each lighting system is presented. II. STRUCTURE OF THE HUMAN EYE AND VISUAL CONDITIONS ON DIFFERENT LEVELS OF LIGHT The eye, or eyeball, is one of the most sophisticated and developed sense organs in humans. This presents a high sensitivity and accuracy in the perception of reflected light by objects around us [6]. Although we can rely on our eyes to bring us most the information from the outside world, they are not able to reveal everything. They can see only objects that emit or are illuminated by light waves in our perception ranging from 380 to 780 nm, approximately [7]. The rods and cones are the photoreceptor cells that compose the sensory layer of the retina, and they are sensitive to visible light stimulation. Cones are cells responsible for chromatics vision and visual acuity, reacting quickly to detect details and colors. This photoreceptor works for daytime vision, which is called photopic vision. The rods are photoreceptor cells responsible for vision in low light conditions, or scotopic vision. They are involved in no chromatics vision. Because of the distinction between rods and cones, different sensitivity curves were defined for both types of photoreceptor cells [8]. III. V AND V CURVES AND THE LUMINOUS EFFICIENCY The human vision pattern has been defined by experiments in the 20s and standardized by the CIE (Commission Internationale de l Eclairage) [9]. These experiments determined the human visual sensitivity light spectral, defining the V curve, which represents a human spectral response under photopic However, there are situations when the high brightness conditions of photopic vision are no longer met. As a result, the V curve was

2 obtained. This curve represents a human spectral response under low light conditions, the scotopic vision. Fig. 1 shows the V and the V curve. According to the V curve, the cones, which are responsible for the photopic vision, have a defined maximum luminous flux of 683 lm. This index, when related to efficiency, corresponds to the maximum luminous efficiency (683 lm/w) that a hypothetical monochromatic radiation light source at 555 nm wavelength can reach. Nevertheless, the rods, which are responsible for the scotopic vision curve V, are more receptive to light at a 508 nm wavelength (scotopic sensitivity peak), with a maximum luminous flux set up in 1,700 lm. From these observations one can define certain light sources lumens [10]. However, there are intermediary lighting conditions in which there is the excitement of both cones and rods. This condition is named mesopic vision [11]. IV. LAMP LUMENS DETERMINATION UNDER PHOTOPIC AND SCOTOPIC CONDITIONS The lamp lumens determination involves knowledge of the spectral power distribution (SPD) of the lamp and the eye visual response. The light is defined as the energy measured by the human eye. Thus, light is not simply defined as energy. When compared to other radiation types, it is defined as the visual effect created by the energy. Moreover, it is recalled that the human eye spectral response, the V curve, is defined under photopic visual conditions and the V curve under scotopic visual Equation (1) is used to determine the light source photopic lumens () [10]. This is the way used by the lamp manufacturers, in order to define the lamp lumens. Therefore, the light sources efficiency is based on the photopic lumens. φ = P = Constant of the maximum light output at 555 nm; P = Power in a certain wavelength; V = Photopic sensitivity curve; d = Wavelength increasing. However, if the visual conditions are changed, the V curve is no longer applicable to determine the lamp lumens, not offering an accurate indication of the luminous flux produced [12]. As a result, in order to set the modified lamp lumen output considering the change of the eye sensitivity in low light levels (scotopic vision); the term "effective lumens" ( ) is used. The light source effective lumens are defined by equation (2) [11]. 780 φ ' = 1,700 P V ' 380 V d d 1,700 = Constant of the maximum light output at 508 nm; V = Scotopic sensitivity curve. (1) (2) Fig. 1. V and V curves of the human eye spectral sensitivity and the values of luminous flux corresponding. V. SPECTRAL ANALYSIS OF ARTIFICIAL LIGHT SOURCES As noted earlier, the lamp lumens value varies with the light source SPD and eye visual response. The following is an analysis of how the four main illumination sources behave: high pressure sodium (HPS) lamp, fluorescent lamp, metal halide (MH) lamp and power LEDs, in photopic and scotopic A. HPS (High Pressure Sodium) Lamp The HPS lamp behavior is based on an electrical discharge in a tube containing sodium and mercury vapor under high pressure, creating an arc discharge within the tube, which gives rise to electrically excited sodium atoms, which emit radiation at wavelengths characteristic of sodium, around 589 nm [13]. The HPS lamp is marketed in a wide power range (35 to 1,000 W) with a lifetime that may vary from 10,000 to 32,000 hours, which means lower maintenance and replacement cost than some lamps, such as fluorescent. Moreover, this lamp has a luminous efficiency up to 150 lm/w, considering the V sensitivity curve. But, it has a low color rendering index (CRI), about 22 %. For these reasons, the HPS lamp is widely used in exteriors lighting systems, for example, in public lighting. Due to the HPS lamps low CRI, others lighting systems, such as metal halide lamps and LEDs, have been research subject for use in public lighting. In [14] and [15] a study about the LEDs use for outside lighting is performed. Fig. 2 (a) shows the HPS lamp (MASTER SON Plus PIA W) SPD. The reason for a high luminous efficiency of HPS lamps is due to the maximum output power located in the yellow region, which has much influence on the human eye when viewed from the curve V. As the lumens are defined by the light amount that the eye perceives under photopic conditions, HPS lamps have a high luminous efficiency. This is because the energy peak emitted by the lamp is very close to the human eye sensitivity peak, in the photopic condition (555 nm). It is also evident that at wavelengths around 508 nm, there is low output power, thus, the effective lumens in the scotopic condition are reduced. The HPS lamp produces very little blue and green light and

3 thus their efficiency under low illumination is significantly reduced. This lamp type efficiency, in scotopic conditions, decreases to about 1.5 times less than the efficiency in photopic This result shows that the HPS lamp model reaches approximately 50,800 effective lumens in photopic B. Fluorescent Lamp These discharge lamps convert about 60 to 70 % of electrical power in ultraviolet radiation (UV) [16]. This UV radiation is converted into visible light by a phosphor layer coating inside the discharge tube. The fluorescent lamp characteristics are directly related to the phosphor layer. If used, for example, rare earth metals called triphosphors, the lamp s CRI can reach 85 %, and efficiency of 100 lm/w [17]. Fig. 2 (b) shows the fluorescent lamp (Model TL W) SPD. For this lamp type, the radiation is presented in almost all visible wavelengths, with peaks in the blue, violet, blue-green and yellow regions. When the output power is compared to the photopic sensitivity curve, the result is a considerable amount of lumens. In scotopic conditions, some fluorescent lamp energy peaks are located near the eye high sensitivity regions, especially in the region of blue and bluegreen. The result is that the fluorescent lamp effective lumens increase as the light level is low, and when the eye response passes from photopic to scotopic. This lamp type has efficiency, in scotopic conditions, approximately 2.0 times higher than the efficiency in photopic This result shows that this fluorescent lamp model reaches approximately 4,320 effective lumens under scotopic C. MH (Metal Halide) Lamp The MH lamp presents an excellent CRI (aproximatelly 90 %) and a luminous efficiency exceeding 100 lm/w [18], with a lifetime ranging between 5,000 and 22,000 hours. Currently, this lamp type is marketed in the power range between 20 and 2,000 W [19]. Fig. 2 (c) shows the MH lamp (HQI T /400 W) SPD. This lamp exhibits a considerable energy output in almost all wavelengths, with peaks in the blue-green region. When the lamp output power is compared to the photopic sensitivity curve, the result is a high lumens output. Besides, considering scotopic conditions, the result is also high effective lumens, because of the light emission in the regions of blue and blue-green. A performance study about the MH lamps has been developed in [20], considering the visual conditions present in a public lighting system. This lamp type has efficiency, in scotopic conditions, approximately 1.7 times higher than the efficiency in photopic This result shows that this MH lamp model reaches approximately 72,400 effective lumens in scotopic From this analysis it is concluded that the efficient lumens increase for MH lamps as the light level is low, in contrast to the HPS lamps. D. LED (Light Emmiting Diode) LED is a semiconductor device that, when forward biased, emits radiation visible or not to the human eye. The electrical energy conversion into light is made on the material, and therefore it is so called solid state lamp. With small size, LED has become an alternative to artificial lighting [21]. The advantage of using LEDs in lighting systems is due to its high luminous efficiency, up to 60 lm / W, and long lifetime, 50,000 to 100,000 hours. In addition, it has a high CRI, between 70 and 80 %. However, the light color emitted by the diode depends on the semiconductor material with which it is built and the phosphor layer used. LED s correlated color temperature can range from 2,700 to 6,000 K, being employed in a wide range of lighting systems [22]. Fig. 2 (d) shows the LED (Model 7007-PWC-10-1/LED 3 W) SPD. It is observed that there is a large power output mainly in the blue (450 nm) and green-yellow (550 nm) regions. This LED type has efficiency, in scotopic conditions, approximately 2.3 times higher than the efficiency in photopic This result shows that this LED model reaches approximately 415 effective lumens in scotopic When the LED s output power is compared to the human eye photopic sensitivity curve, there is a high lumens output near the eye sensitivity peak, a factor that provides a high luminous efficiency under high lighting. The same is observed for the scotopic response, due to the blue range emission. VI. PHOTOPIC AND SCOTOPIC COMPARATIVE STUDY AMONG THE PROPOSED LIGHT SOURCES The Lawrence Berkeley National Laboratory carried out measurements of luminous efficiency of HPS lamp, fluorescent lamp and MH lamp [23]. In [14] the analysis of luminous efficiency of LEDs was carried out. Both these measurements were taken at different brightness levels, so that the luminous efficiency results for photopic vision and the scotopic vision can be reached. Table I provides a comparison among the studied models of HPS lamp, fluorescent lamp, MH lamp and LED, in photopic and scotopic From the analysis of the Table I, one can establish a comparison degree among different light sources luminous efficiency, both in terms of photopic and scotopic visions. It is observed that there is a great variation of the fluorescent lamp luminous efficiency, for example, between the visual VII. STUDY CASE Lighting systems are responsible for a great amount of the electrical energy consumed worldwide. This index involves a series of research projects to the illumination area to provide more efficient systems, both in terms of light emitted and power consumption. This paper proposes a public road lighting study case, such as a crossroad lighting, in order to evaluate the best suited light source for this environment. In this study, luminous efficiency and the power consumed are presented, as well as the system implementation cost, under photopic and scotopic

4 Fig. 2. Lamps spectral distribution. A. Case proposed A crossroad lighting This section presents a crossroad illumination case study, which can represent photopic or scotopic conditions depending on the light level, in order to evaluate the proposed analysis. The crossroad illumination is designed to provide an environment with a well distributed light, offering security for the vehicles and even pedestrians passage. Durability, performance, high efficiency, and low implementation cost are the main lighting system features for such an environment. Since the work objective is not to determine the luminous flux level required, it was adopted a luminous flux of 203,200 lumens for calculation. This value corresponds to the luminous flux of four 400 W HPS lamps, under photopic conditions, for a hypothetic crossroad lighting. B. Number of lamps required under photopic and scotopic conditions For the lighting of the proposed environment, it has been established the possibility of using four types of lighting sources: HPS lamp, fluorescent lamp, MH lamp and LEDs. Table II shows the characteristics of the studied lamp types with their luminous flux variations in different lighting Table III shows the number of lamps necessary to achieve the luminous flux of the proposed case. According to the Table III analysis, it can be observed that, depending on the light situation in the environment (photopic or scotopic), the number of lamps to be used can be reduced without human eye performance degradation. C. Power consumption according to the lamps number Table IV shows the rates of electric power consumed by HPS lamp, fluorescent lamp, MH lamp and LED, according to the number of lamps used in photopic and scotopic conditions (Table III), to illuminate the proposed environment. According to the Table IV analysis, it can be observed that, in photopic conditions, the HPS lamps are the most efficient light source. However, when considered in the scotopic perspective, the MH lamps consume less than a half of the electric power consumed by the HPS, moreover, at this same visual condition, the MH lamps have the highest luminous efficiency. Considering the power levels required for each lamp type, to illuminate the proposed environment, it is observed that in matters of lighting systems efficiency improvement, it would be more effective the use of lamp types with different power spectra, according to the environment light condition. D. Energy cost in different visual conditions The consumed electricity cost is one of the relevant factors taken into account in the lamp type definition, to be used for illuminating certain environments. Table V provides a link between each lamp power consumption and these consumption costs. It was considered a lamps operation period of 12 hours a day for 30 days at a rate of US$ 0.18 / kwh. TABLE I PHOTOPIC AND SCOTOPIC LUMINOUS EFFICIENCY. Light source Photopic (lm/w) Scotopic (lm/w) Fluorescent HPS MH LED TABLE II LAMPS FEATURES UNDER DIFFERENT LIGHT LEVELS. Light source Power (W) Photopic (lm) Scotopic(lm) Fluorescent 40 2,180 4,320 HPS ,800 32,240 MH ,800 72,400 LED TABLE III TOTAL LAMPS NUMBER IN DIFFERENT VISUAL CONDITIONS. Light source Photopic vision Lamps Scotopic vision Lamps Number Number Fluorescent HPS 4 7 MH 5 3 LED 1, TABLE IV TOTAL POWER UNDER PHOTOPIC AND SCOTOPIC CONDITIONS. Light source Power consumed in Power consumed in photopic conditions scotopic conditions Fluorescent 3,760 W 1,880 W HPS 1,600 W 2,800 W MH 2,000 W 1,200 W LED 3,387 W 1,473 W

5 TABLE V ENERGY COST IN PHOTOPIC AND SCOTOPIC CONDITIONS FOR 30 DAYS. Light source Monthly cost in photopic Monthly cost in scotopic conditions conditions Fluorescent HPS MH LED It is concluded that the fluorescent lamps use under photopic conditions would result in a high cost, about US$ Now, with the HPS lamps use, the energy cost would be reduced to less than half. However, when considered scotopic conditions, it appears that the HPS lamps have a cost around 1.48 times higher than the fluorescent lamps cost under the same Metal halide lamps use under photopic conditions would result in an intermediary cost, about US$ However, when considered scotopic conditions, it appears that the metal halide lamps cost would be around 1.68 times lower than the same lamp in photopic conditions, which results the lowest energy cost among all studied lamps. In addition to the energy cost, the system maintenance and lamp lifetime are relevant factors that have much influence in the overall cost of the system. However, these issues are not covered in the paper focus. E. Lighting system investment cost The light system investment cost is directly related to the amount of lamps, which changes according to the visual From this variation, one can estimate an approximate investment value needed for the lighting system implementation. Fig. 3 shows the approximate investment cost for the studied lamps in both visual For the calculation criteria, it was determined the 40 W fluorescent lamp with ballast cost at US$ The 400 W HPS lamp with ballast cost at US$ The 400 W MH lamp with ballast cost at US$ Yet the 3 W LED system cost at US$ 7.6 the unit. According to the Fig. 3 analysis, the illumination with high pressure sodium lamps is around 3.35 times cheaper when compared to fluorescent lamps in photopic condition. However, when these values are analyzed in scotopic visual conditions, both systems implementation prices difference are reduced considerably, becoming almost negligible. It is observed that the LEDs use in the lighting system proposed has the highest initial investment cost due to the large components number. An overall cost analysis, neglecting the maintenance cost, can be conducted over a longer time period, in order to evaluate the lighting system benefit cost. Given each lamp type monthly cost values under photopic and scotopic conditions, obtained in Table V, and the investment cost presented in the Fig. 3, the overall cost through a 10 years period is shown in Fig. 4. According to the Fig. 4 analysis, it is observed that, initially, considering a photopic vision condition, the HPS lighting system has the lowest cost. However, in scotopic vision conditions the system that has the lowest implementation cost over 10 years was based on MH lamps. Fig. 3. Lighting system investment value required for both lamps. Fig. 4. Overall cost in a 10 years period under photopic and scotopic visual conditions

6 However, in a lighting project, the initial investment is not the only relevant factor that may determine the lamp type to be used for environment lighting. The lifetime is an important factor in choosing an adequate lighting system. Whereas a time span of 10 years, with a total of 43,800 hours of system operation, the only light sources that do not need to be replaced are the LEDs lamps. VIII. CONCLUSION The various artificial lighting sources that are currently marketed can be compared to each other using various criteria such as price, lifetime, and color rendering index. However, a very important criterion is the luminous efficiency (lumens/watt), or how much light is emitted per electrical power unit, the latter being directly related to energy efficiency improvements. Recent literature indicates that the light source spectral distribution has an effect on the visibility that it produces. A high light intensity excites eye photoreceptors that are sensitive to certain wavelengths, the cones. However, in low levels of light, the excitement of other photoreceptor cells occurs, the rods. They are responsible for scotopic vision, which has different sensitivity response when compared to photopic vision. As a result, it is concluded that using lamps with the adequate light spectrum, light levels can be reduced without visual performance compromising. This implies a considerable lighting systems total power used reduction, without any change in relation to human eye. In the examined street lighting study case, one can conclude that the power variation required to maintain the same luminous flux, is considerably high in the different visual On this basis, it appears that a lighting design project that aims an efficient lighting should consider such power consumption changes between the photopic and scotopic vision, in order to achieve the lowest power consumption. For the case study conducted, in which the goal is to illuminate a crossroad, it is concluded that for photopic conditions it would be more advantageous to use HPS lamps. However, the crossroad lighting environment can be suited as scotopic vision, in which is more profitable the use of MH lamps. ACKNOWLEDGEMENT The authors would like to acknowledge the Federal University of Santa Maria and CNPq for funding (478676/2009-3), and the GEDRE - Group of Intelligence in Lighting for the technical support. REFERENCES [1] Wu, J., H. Kita and Y. Nishikawa (1993). A four-layer neural network model of the equivalent luminous-efficiency function in the human vision. IEEE International Joint Conference on Neural Networks, vol. 1, [2] Jones, B. F. and P.E., F.I.E.S (1989). The influence of spectral energy distribution of light sources on visual performance. IEEE Industry Applications Society Annual Meeting, vol. 2, [3] Kostic, M. and L. Djokic (2009). Recommendations for energy efficient and visually acceptable street lighting. 11th Conference on Process Integration, Modelling and Optimisation for Energy Saving and Pollution Reduction (October), vol. 34, Issue 10, [4] Josefowicz, J. and D. Ha (2008). Vision and Exterior Lighting: Shining Some Light on Scotopic and Photopic Lumens in Roadway Conditions. Led Roadway Lighting, [5] Akashi, Y., M. Pete and M. Rea (2006). An energy-efficient street lighting demonstration based upon the unified system of photometry. Lighting Research Center, Rensselaer Polytechnic Institute (april), 1-6. [6] Land, M. F. and R. D. Fernald (1992). The evolution of eyes. Annual review of neuroscience (March), vol. 15, [7] Bullough, J. D., and M. S. Rea (2004). Visual performance under Mesopic Conditions. Journal of the Transportation Research Board (August), [8] Damelincourt, J. J. (2000). Lamps and lighting. IEEE Engineering Science and Education Journal (October), vol. 9, [9] Viikari, M., M. Eloholma and L. Halonen (2005). 80 years of V() use: a review. Light & Engineering, vol. 13, [10] Borton, J. A. (1997) A comparison of light sources for the petrochemical industry. IEEE Industry Applications Magazine (July August), vol. 3, [11] Walkey, H. C., J. A. Harlow and J. L. Barbur (2006). Characterising mesopic spectral sensitivity from reaction times. Applied Vision Research Centre, The Henry Wellcome Laboratories for Vision Sciences (October), vol. 46, [12] Lewin, I. (1999). Lamp color and visibility in outdoor lighting design. Conference of the Institution of Lighting Engineers Portsmounth, England. [13] Groot, J. and J. Van Vliet (1986). The High-Pressure Sodium Lamp. 1st Edition. London: Macmillan Educational. [14] Dalla Costa, M. A., G. H. Costa, A. S. Santos, L. Schuch and J. R. Pinheiro (2009a). A high efficiency autonomous street lighting system based on solar energy and leds. The 10th Brazilian Power Electronics Conference. [15] Long, X. and J. Zhou (2008). An Intelligent Driver for Light Emitting Diode Street Lighting. IEEE Automation Congress (Sept - Oct) [16] Hammer, E. E. (1983). Fluorescent Lamp Starting Voltage Relationships at 60 Hz and High Frequency. IEEE Journal of the Illuminating Engineering Society (October), vol. p [17] Meyer, C. and H. Nienhuis (2008). Discharge lamps. Philips Technical Library (Kluwer Technisch Boeken BV). [18] Sugiura, M. (1993). Review of metal-halide discharge-lamp development Science, Measurement and Technology, IEEE Proceedings A (November). vol. 140, [19] Waymouth, J. F. (2005). Metal Halide Lamps. Proceedings of the IEEE Publication, vol. 59, [20] Dalla Costa, M. A., M. S. D. Perdigão, H. A. C. Braga, J. M. Alonso (2009b). Projeto e implementação de um reator eletrônico microcontrolado, de elevado fator de potência, orientado a lâmpada de vapor metálico. SBA Controle & Automação (September), vol. 20, [21] Crawford, M. H. (2009). LEDs for Solid-State Lighting: Performance Challenges and Recent Advances. Selected Topics in Quantum Electronics IEEE Journal (July-august), vol. 15, [22] Luxeon (2008). Power light source Luxeon Rebel: LXML-PWN Spec Sheet. [23] Rea, M. S. (1995). In The Dark About The Lumen. IAEEL newsletter, Lighting Research Center Rensselaer Polytechnic Institute.