Artificial lighting in office buildings typically requires 30% of the total energy consumption of the building, providing a substantial opportunity for energy savings. To reduce the energy consumed by indoor lighting, we propose a sensor-less light-emitting diode (LED) dimming system using daylight harvesting. In this study, we used light simulation software to quantify and visualize daylight, and analyzed the correlation between photovoltaic (PV) power generation and indoor illumination in an office with an integrated PV system. In addition, we calculated the distribution of daylight illumination into the office and dimming ratios for the individual control of LED lights. Also, we were able directly to use the electric power generated by PV system. As a result, power consumption for electric lighting was reduced by 40 – 70% depending on the season and the weather conditions. Thus, the dimming system proposed in this study can be used to control electric lighting to reduce energy use cost-effectively and simply.
© 2013 Optical Society of America
To reduce carbon emissions and promote green development, the need to reduce energy use in buildings has gradually increased [1–3]. Because 30% of the total energy use in buildings is for electric lighting, this area is an important target for reduction so that we are willing to exchange the present electric lightings with energy favorite LED lightings [4–6]. Also, there has been increasing interest in indoor lighting using daylight, because daylight comes from renewable solar energy and can reduce energy use and provide environmentally friendly lighting [7–11].
In general, indoor lighting is maintained at a constant illumination although the brightness of the external environment changes, using unnecessary electric energy. Many approaches have been attempted to reduce this waste energy in buildings in various fields of the lighting industry [12–15]. Among these approaches are widely used methods for reducing artificial light and saving energy when daylight reaches the windows of buildings, known as daylight harvesting [16–20]. Other popular methods include the use of on/off lighting controls that sense occupancy, and dimming controls that can supply the proper illumination to interior space and reduce excess energy use . However, for both of these methods, many illuminance sensors are needed to control lighting and supply the proper illumination. These illuminance sensors are typically fabricated with cadmium sulfide (CdS), require an additional microcontroller unit (MCU), and the non-linear characteristics of the sensors result in large tolerances, although they are inexpensive and small. Photodiodes fabricated using semiconducting processes are small and have rapid response to changes in light. Nevertheless, these two types of sensors can be greatly influenced by movements of occupants and may malfunction. Many sensors are also needed inside and outside offices or rooms [22–24].
The other effort to reduce the energy use for indoor lighting is to use the photovoltaic system installed outside or on the roof of the office building . This method is good to supply the electric power to operate the indoor lighting, but it should need additional batteries. And, in order to use the electric power generated by the photovoltaic system, the conversion process is needed, including converting DC to AC and AC to DC. These conversion processes could be one of reasons to reduce the efficiency of photovoltaic-based electric power supply system .
Therefore, in this study, we introduce a new approach to controlling electric lighting using daylight harvesting through photovoltaic power generation in office buildings with building-integrated photovoltaic (BIPV) systems. The daylight distribution after transmitting through a window shows the exponential decay along the distance from the window so that we can expect the daylight distribution in the office. Also, the amount of daylight coming into the office are closely related to the amount of electric power generated by BIPV. Thus, without any illuminance sensors or photodiodes, artificial lighting in offices can be controlled through prediction of daylight inflow to maximize energy savings. In addition, by directly using the electric power generated by BIPV system, we can also maximize energy savings for indoor lighting because there is no conversion loss of converting DC to AC and AC to DC. For these purposes, we used energy-friendly light-emitting diodes (LED) using the power generated by BIPV systems. To confirm the effectiveness of our new approach, we simulated the dimming control system in a pilot test and calculated energy reduction and the resultant energy savings.
In an office building using photovoltaic power generation, we proposed an alternative artificial lighting control system based on dimming control without additional photo-sensors. As shown in Fig. 1, our novel method is specially designed for office buildings facing to south with integrated photovoltaic power generation systems. Based on the amount of power generated by the Building Integrated Photovoltaic (BIPV) system, we can determine the level of daylight passing through the windows, the daylight distribution inside the office, and calculate the dimming ratio according to our proposed algorithm without the need for photo-sensors in order to satisfy the minimum indoor illumination.
Generally, the levels of daylight reaching at the surface of window are varying with respect to the external weather conditions, Clear sky or Overcast sky, and also analogue to the levels at the surface of solar panel. The amount of daylight in clear sky is explicitly larger than that of daylight in overcast sky, but after reaching at the surface of window, the daylight distribution decreases exponentially with respect to the distance from the window . Therefore, we can predict the daylight distribution in the office only with the amount of daylight at the surface of window. In addition, the amount of daylight at the surface of window can be expected by the amount of electric power generated by BIPV system because the amount of daylight outside is closely related to generating electric power of BIPV system. Furthermore, the electric power generated by BIPV system can be used to operate the light-emitting diode (LED) lights without any conversion processes, leading to improving the energy savings in terms of energy efficiency.
2.1 Location and experimental configuration
The office building modeled and examined in this study was located in central South Korea, specifically in the city of Daejeon. An office building with three floors was used as the research building facing to south, shown in Figs. 2(a) and 2(b). The longitude and the latitude of this building were 36.22°N and 127.22°E. The dimensions of the test office were 4 m wide, 6.5 m long and 2.7 m high.
2.2 Photovoltaic power generation and lighting system
We installed photovoltaic power generation consisting of 60 W single-crystal silicon solar cells integrated on the wall of the office building. We used three LED lights in the test office with 52 W rated-power, 48 V differential output voltage, 4000 lm luminous flux, 5700 K color temperature, and more than 77 lm/W of luminous efficiency as shown in Table 1. LED lights have advantages for this light dimming system due to the linearity of their optical power and illumination , leading to controlling accurately and maximizing the energy savings.
2.3 Correlation between indoor illumination and external photovoltaic power generation
In order to obtain the daylight distribution in the office, we first measured the indoor illuminations at different positions (1-6 m from the window) due to the daylight. Figure 3 shows the distribution trend of daylight along the distance from the window.
After measuring the daylight distribution, we obtained that the daylight after passing through the window decreased exponentially in Eq. (1).
In Eq. (1), the daylight distribution denoted by y was plotted along the distance from the window, and the constants can be defined as a decreasing factor of daylight along the distance of x. α1 and α2 are dimension-less structural factors coming from the dimension of office. These two structural factor can be obtained simply by experimental measurement with respect to the office structure. Especially, α1 is related to window to wall (WW) ratio. This ratio means the initial value of daylight at the time of passing through the window with respect to the daylight outside. Also, α2 is related to a distance from the window to the wall inside office. In this study, the long of office is 6.5 m from the window.
To obtain the correlation between the distribution of daylight in the office and the electric power generated by the BIPV, we monitored indoor illumination during working hours from Feb. 27, 2012 to Mar. 1, 2012. Indoor illumination was measured using an illuminometer 0.75 m above the floor and 6 m from the window as shown in Fig. 4.
Over the working hours measured, photovoltaic power generation varied with the change in indoor illumination from the window. The weather was clear sky (parabolic curves) on Feb. 27, Feb. 29, and Mar. 1 and overcast sky (non-parabolic curve) on Feb. 29. We then calculated the absolute ratio of indoor illumination to the amount of photovoltaic power generation to obtain the conversion factor as shown in Fig. 5, which averaged 4.24. This conversion factor can allow the prediction of illumination one-dimensionally in the office, e.g., at 1, 2, 3, 4, and 5 m, and it is described as Eq. (2).
For the experimental test of target office in this study, we obtained a conversion factor β of 4.24 and structural factors α1 and α2 of 45.11 and 0.84, respectively. The structural factors vary with the dimensions of the office, the type, direction, and reflectance of the windows, etc. Based on Eq. (2) representing the correlation between photovoltaic power generation and indoor illumination, we were able to predict the indoor illumination from daylight at each point in the office.
2.4 Calculation of the dimming ratio
Using the correlation between photovoltaic power generation and indoor illumination at 6 m in the office, we first obtained the equation for indoor illumination due to the daylight distribution in Eq. (2). To calculate the dimming ratio for each of the three LED lights, we obtained the input signal P, which is a power generated by BIPV. And, we also calculated the indoor illumination 0.75 m above the floor at each point corresponding to the lights, i.e., 1 m, 3.7 m, and 6 m away from the window. The expected indoor illumination at each point, Lx, can be obtained using Eq. (3):
Once the indoor illumination at each point was determined, we calculated the dimming ratio for each of the LED lightings by using Eq. (4):
If the expected indoor illumination (Lx) was more than recommended illumination (LR), especially 500 lx in South Korea, then we set the dimming ratio to 0% because the minimum requirement for light was satisfied . In addition, if the dimming ratio was more than 0% and less than 10%, we could not control the LED lights linearly in the range of 0% to 10% so that the dimming ratio was set to 10% to avoid flickering of the LED lights at low levels. After calculating the dimming ratio for each LED light, we used pulse-width modulated (PWM) dimming method that has the essential advantage, arising from the constant peak of wavelength in PWM dimming control.
Based on the simulation results, we tested the dimming control system described in Sec. 2 with three LED lights located on 1 m, 3.7 m and 6 m away from the window. The dimming control system consisted of three parts shown in Fig. 6.A 60 W solar cell array as shown in Fig. 6(a) was installed outside the test office, and three LED panel lights as shown in Fig. 6(c) were installed inside the test office. Figure 6(b) shows measurement and control system. At first, we measured solar power generation a digital signal processor (DSP TMS320F28335; Texas Instruments), and calculated the indoor illumination and dimming ratio based on Eqs. (1)–(4). The calculated dimming ratio was used as the PWM dimming input for the LED driver to individually control the three LED lights. This experiment was carried out under two weather conditions, clear sky and overcast sky, for both the simulation and pilot test. Finally we obtained the expected energy savings under each weather condition.
A circuit diagram for the LED light dimming control system is shown in Fig. 7. The electric power generated by the BIPV was used as an input signal to calculate the amount of daylight in the office and the dimming ratio for each LED light. The electric power of the BIPV can be also used to supply directly the electric power needed to operate the LED lights when daylight is insufficient for indoor illumination. The DSP was used to detect the voltage, current, and electric power generated by the BIPV and calculate the need for additional illumination at each point in the office based on the generated electric power. Thus, the three LED lights were individually controlled to meet the recommended illumination of 500 lx (in the case of South Korea) in the office. In addition, the power supply was designed to directly use the DC power generated by the BIPV system, leading to reducing the conversion loss from DC to AC through a DC/AC converter. In general, in order to operate the LED lights by using the photovoltaic system, we should perform to convert DC to AC and AC to DC so that the conversion loss increases through each step. However, we can transmit directly DC power generated by the BIPV system to each LED light without any conversion. Thus, we can also increase the efficiency of power transmission in this study.
4.1 Simulation results
We first confirmed our dimming system using PSIM simulation software (Powersim, Inc.) based on Eqs. (1)–(4) and the dimming ratio described in Sec. 3.2. We calculated the dimming ratio at each point in time for both clear and overcast sky conditions. We also predicted the energy savings for each sky condition. Figure 8 shows the indoor illumination at each point in time and the power consumption.
The simulation results for indoor illumination shown in Figs. 8(a) and 8(c) demonstrate the ability of this system to satisfy the minimum requirements for indoor illumination of 500 lx. Figures 8(b) and 8(d) show the power consumption of the electric lights with daylight harvesting. The three LED lights used 156 W/h, a total of 1404 W during the working day without dimming control. When dimming control was conducted under clear skies, 813 W/day were used, reducing the electric energy consumption by ~42%. Additional renewable solar energy was used because the 60-W solar panel arrays generated 408 W over 9 h; thus, the electrical energy used was reduced by 71% on a clear day. Energy savings for overcast skies were less than for clear skies, although the minimum requirement for indoor illumination was still satisfied, as shown in Figs. 8(c) and 8(d). Under overcast skies, power consumptions was also 1404 W/day with no dimming control, and was 1178 W/day with dimming control. Due to the cloudy weather, the solar power system generated only 77 W, less than under clear skies, so that energy consumption by the indoor lights was reduced by 16% with dimming control only and 21% for dimming control with the added use of generated solar power.
4.2 Experimental results
After confirming the operation of our dimming control system with the simulation results, we tested our dimming control system in an actual office. Figure 9 shows the measured indoor illumination due to only daylight distribution and indoor lightings with daylight. Solid line and dotted line in Fig. 9 plotted daylight distribution and daylight distribution with indoor lightings under clear and overcast skies.
In Fig. 9, under clear and overcast skies, the indoor illuminations with artificial indoor lightings satisfied the recommended minimum indoor illumination of 500 lx in all times and on all positions. However, without artificial indoor lightings, only the positions of 1 m and 2 m satisfied 500 lx. Therefore, in order to satisfy the recommended indoor illumination, the artificial indoor lightings should be adjusted by dimming control for satisfying the recommended indoor illumination and saving the energy use of artificial lightings.
Figure 10 shows the measured indoor illumination and associated power consumption. The measured indoor illumination was influenced by the daylight distribution in the office and the artificial LED lights controlled by the dimming ratio. The experiments were performed under clear skies (30 Apr 2012) and overcast skies (May 1, 2012).
As shown in Figs. 10(a)–10(c), in all areas of the office, indoor illumination satisfied the minimum requirement of 500 lx. Figures 10(b) and 10(d) show the power consumption of the electric lights with daylight harvesting. Under clear skies (Apr 30, 2012), the three LED lights used 156 W/h, a total of 1404 W over the working day without dimming control. When dimming control was added, 685.2 W/day was used, reducing the electricity consumption by ~51%. Additional renewable solar energy of 361 W/day was generated by the 60-W solar cell arrays over 9 h. Thus, electrical energy use was reduced by up to 75% on a clear day. As expected, overcast conditions (May 1, 2012) resulted in less energy savings, while still satisfying the minimum requirement for indoor illumination, as shown in Figs. 8(c) and 8(d). Under overcast skies, the power consumption of the lights was 1404 W/day without dimming control and 822.7 W/day with dimming control. Due to the cloudy weather, the solar power system generated 206.2 W. Therefore, electrical energy use for the indoor lights was reduced by up to 41% with dimming control and 56% for dimming control with added generated solar power.
During the dimming control test of the three LED lights, we monitored the change in the dimming ratios with solar power generation as shown in Fig. 11.LED1, LED2, and LED3 were installed at distances of 1 m, 3.75 m, and 6 m from the window, respectively. As shown in Fig. 11(a), under clear conditions, the closest light (LED1) to the window was fully influenced by daylight during working hours and did not need to be controlled. The dimming ratios for LED2 and LED3 changed in response to the amount of daylight. In addition, under overcast weather conditions, daylight rapidly decreased with time after 3:00 PM, and the dimming ratio for LED1 increased simultaneously with those for LED2 and LED3 in Fig. 11(b).
4.3 Expected energy savings
The annual energy savings can be obtained by averaging energy savings of each day in a year. We monitored the BIPV power generation for a year, Jan 1, 2012 to Dec 31, 2012. Based on the amount of BIPV power generation for each day, we estimated the energy savings of each day and averaged the annual energy savings using our proposed dimming method. As shown in Fig. 12, the average energy savings were estimated by ~59.81% in a year. For January, November, and December, the energy savings of each case were under the average annual energy savings due to low meridian transit altitude. During this period, the power generation of BIPV was slightly less than that of spring and fall seasons so that the amount of electric power used for LED lights was small. Therefore, the effect on energy savings was reduced. Also, the typical weather conditions in July and August are the rainy season in South Korea. Due to this seasonal condition, the average energy savings in July and August also were slightly reduced in comparison to the average annual energy savings.
In this study, we propose a novel approach to controlling electric indoor lighting such as LED lights using a photovoltaic system integrated into an office building. This research suggests two credential points: (1) Based on the amount of available daylight, we can calculate the amount of artificial illumination required with the reference indoor illumination, and we can also control the indoor lightings without the need for indoor photo-sensors. (2) Building integrated PV system can supply the electric power to operate the indoor lightings without conversion loss by supplying direct current (DC) power to LED lightings. Before we applied our dimming system to an actual office, we obtained a correlation between photovoltaic power generation and indoor illumination by daylight. We then confirmed the ability to control indoor lighting using only photovoltaic generation, without any photo-sensors in the office. Based on the predicted distribution of indoor daylight based on outdoor photovoltaic power generation, we calculated the requirement for additional illumination to meet the minimum value of recommended illumination, especially 500 lx in South Korea, and controlled each LED light to provide this illumination. The results of the simulation and pilot test experiment showed indoor illumination from daylight and the artificial lights using the dimming system was able to meet the minimum requirement of 500 lx. Particularly in office buildings with an integrated photovoltaic system, we anticipate energy savings of 57-75% under clear skies and 13-29% under overcast skies. Expected annual energy savings are approximately 59.81%.
This work, as a core research project, was financially supported by the Korea Institute of Energy Research and a National Research Foundation of Korea (NRF) grant funded by the Korea Government (MEST, No. 2011-002805).
References and links
1. D. Chwieduk, “Towards sustainable-energy buildings,” Appl. Energy 76(1-3), 211–217 (2003). [CrossRef]
2. A. M. Omer, “Energy, environment and sustainable development,” Renew. Sustain. Energy Rev. 12(9), 129–163 (2008). [CrossRef]
3. L. D. D. Harvey, “Reducing energy use in the building sector: measures, costs, and examples,” Energy Effic. 12, 2265–2300 (2009).
4. C.-H. Tsuei, J.-W. Pen, and W.-S. Sun, “Simulating the illuminance and the efficiency of the LED and fluorescent lights used in indoor lighting design,” Opt. Express 16(23), 18692–18701 (2008). [CrossRef] [PubMed]
5. C.-H. Tsuei, W.-S. Sun, and C.-C. Kuo, “Hybrid sunlight/LED illumination and renewable solar energy saving concepts for indoor lighting,” Opt. Express 18(S4Suppl 4), A640–A653 (2010). [CrossRef] [PubMed]
6. S. H. Lee and J. K. Kwon, “Distributed dimming control for LED lighting,” Opt. Express 21(S6), A917–A932 (2013). [CrossRef]
7. D. H. W. Li and J. C. Lam, “Evaluation of lighting performance in office buildings with daylighting controls,” Energy Build. 33(8), 793–803 (2001). [CrossRef]
8. W. Maranda and M. Piotrowicz, “Application of photovoltaics for daytime indoor lighting,” in Mixed Design of Integrated Circuits and Systems (MIXDES),Proceedings of the 18th International Conference (2011), pp. 529–532.
9. T. Miyazaki, A. Akisawa, and T. Kashiwagi, “Energy savings of office buildings by the use of semi-transparent solar cells for windows,” Renew. Energy 30(3), 281–304 (2005). [CrossRef]
10. D. H. Li, T. N. Lam, W. W. Chan, and A. H. Mak, “Energy and cost analysis of semi-transparent photovoltaic in office buildings,” Appl. Energy 86(5), 722–729 (2009). [CrossRef]
11. K. Kapsis and A. K. Athienitis, “Building integrated semi-transparent photovoltaics: energy and daylighting performance,” Proc. SPIE 8007, 800726 (2011).
12. F. Rubinstein, M. Siminovitch, and R. Verderber, “Fifty percent energy savings with automatic lighting controls,” IEEE Trans. Ind. Appl. 29(4), 768–773 (1993). [CrossRef]
13. G. F. Min, E. Mills, and Q. Zhang, “Energy efficient lighting in China: Problems and prospects,” Energy Policy 25(1), 77–83 (1997). [CrossRef]
14. B. Cook, “High efficiency lighting in industry and commercial buildings,” Power Eng. J. 12(5), 197–206 (1998). [CrossRef]
15. T. Taguchi, “Present status of energy saving technologies and future prospect in white LED lighting,” IEEJ Trans. Electr. Electron. Eng. 3(1), 21–26 (2008). [CrossRef]
16. R. P. Leslie, R. Raghavan, O. Howlett, and C. Eaton, “The potential of simplified concepts for daylight harvesting,” Lighting Res. Tech. 37(1), 21–40 (2005). [CrossRef]
17. J. T. Kim and G. Kim, “Overview and new developments in optical daylighting systems for building a healthy indoor environment,” Build. Environ. 45(2), 256–269 (2010). [CrossRef]
18. M. B. C. Aries and G. R. Newsham, “Effect of daylight saving time on lighting energy use: A literature review,” Energy Policy 36(6), 1858–1866 (2008). [CrossRef]
19. C. P. Kurian, R. S. Aithal, J. Bhat, and V. I. George, “Robust control and optimisation of energy consumption in daylight—artificial light integrated schemes,” Lighting Res. Tech. 40(1), 7–24 (2008). [CrossRef]
20. G. R. Newsham, M. B. C. Aries, S. Mancini, and G. Faye, “Individual control of electric lighting in a daylight space,” Lighting Res. Tech. 40(1), 25–41 (2008). [CrossRef]
21. M. Warren, S. Selkowitz, O. Morse, C. Benton, and J. E. Jewell, “Lighting system performance in an innovative daylighted structure: an instrumented study,” in Proceedings of the 2nd International Daylighting Conference, Long Beach (1986), pp. 21–221.
22. F. Rubinstein, “Photoelectric control of equi-illumination lighting systems,” Energy Build. 6(2), 141–150 (1984). [CrossRef]
23. A.-S. Choi, K.-D. Song, and Y.-S. Kim, “The characteristics of photosensors and electronic dimming ballasts in daylight responsive dimming systems,” Build. Environ. 40(1), 39–50 (2005). [CrossRef]
24. E. S. Lee, D. L. DiBartolomeo, and S. E. Selkowitz, “Daylighting control performance of a thin-film ceramic electrochromic window: Field study results,” Energy Build. 38(1), 30–44 (2006). [CrossRef]
25. H. Yang, G. Zheng, C. Lou, D. An, and J. Burnett, “Grid-connected building-integrated photovoltaics: a Hong Kong case study,” Sol. Energy 76(1-3), 55–59 (2004). [CrossRef]
26. U. Herrmann, H. G. Langer, and H. van der Broeck, “Low cost DC to AC converter for photovoltaic power conversion in residential applications,” in 24th Annual IEEE Power Electronics Specialists Conference (1993), pp. 588–594. [CrossRef]
27. D. H. W. Li, G. H. W. Cheung, and C. C. S. Lau, “A simplified procedure for determining indoor daylight illuminance using daylight coefficient concept,” Build. Environ. 41(5), 578–589 (2006). [CrossRef]
28. L. Svilainis, “LED PWM dimming linearity investigation,” Displays 29(3), 243–249 (2008). [CrossRef]
29. Korean Standard Information Centre, http://www.standard.go.kr. Accessed on November 24, 2013.