IMPROVING VISUAL COMFORT AND ENERGY

(1)

Master thesis, 30 hp

MSc programme in Energy Engineering, 300 hp

IMPROVING VISUAL COMFORT AND ENERGY

EFFICIENCY IN A CLASSROOM

A comparative approach of evaluating a lighting design technique and a light sensor

positioning method

Anton Sanaei

(2)

Abstract

The reduction of energy demand and increased energy efficiency is believed to play a major role in tackling the global climate change. Artificial lighting systems in commercial buildings accounts for a substantial proportion of the total energy consumption. Studies shows that significant energy savings can be achieved by improving the energy efficiency with the application of control systems and daylight sensors. However, this may not come to the expense of impaired visual comfort. This study contains a comparative analysis of methods to improve the visual comfort, optimise light sensor placement for proper performance and estimate the potential energy savings for a classroom located in Umeå, Sweden. The term visual comfort and corresponding photometric properties has been evaluated in regards to international standards and recommendations. The lumen method, a lighting design technique, showed promising accuracy in determining a more optimal lighting design, this was confirmed by using the lighting simulation software Dialux evo. The results obtained by the simulation software showed improved visual comfort in terms of illuminance-based metrics that fulfilled the international requirements EN12464-1.

A sensor positioning method denoted as "Point-by-point" has been evaluated for two different lighting designs. The current lighting design is based on the existing lighting arrangement in the classroom, denoted as "current design". The other lighting design is based on improved positioning attained by the lumen method, denoted "Test design 2". The point-by-point method showed considerate accuracy in comparison to reference values obtained by simulations, however the credibility of the method is dependent on the lighting design. The average deviation for the current design were determined to be 23.7 lx (15.16 %), whilst the improved Test design 2 attained an lower average deviation of 8.3 lx (9.20 %). Lighting characteristics of the luminaries also has an impact on the credibility of this method, as uniform lighting proves to be more suitable than non- uniform lighting.

The integration of daylight data in the simulations showed different optimal position for light sensors due to the natural changes in illuminance. Thus, the positions with the most established linear relationships between the light levels on the workplane and ceiling throughout the year were consider to be the most suitable. The point-by-point method for Test design 2 acquired an average deviation of 13.1 lx (16.40 %) in comparison to the daylight simulations. The results showcased that this method may be applied in similar studies in the future. The daylight simulations indicated significant energy savings throughout the lifetime of the lamp. The most substantial savings were obtained in the month of May during the first year (63.4 %). This study demonstrates that daylight harvesting is beneficial despite the latitude of the location with proper dimensioning of the fluorescent lighting system.

(3)

Acknowledgement

First and foremost, I would like to thank my supervisor at Umeå University, Gireesh Nair, for his guidance and support through this project. I would also like to thank Tomas Andersson at Akademiska hus and Shoaib Azizi at Umeå University for providing me with valuable material that improved this work’s quality. Lastly, I would also like to acknowledge the project "Energy efficiency improvements in HVAC of university buildings by studying the occupancy pattern" by the Swedish Energy Agency as the data provided to me was available through this project.

(4)

1 Introduction

Buildings make up for about 31 % of the global final energy demand and accounts for approximately 19 % of all global greenhouse gases [1]. Studies shows that the electrical energy consumption for commercial buildings makes up 25-40 % of the total energy demand [2]. By applying control systems and daylight sensors, its possible to reduce the energy consumption for unnecessary lightning [1]. Daylight harvesting takes advantage of utilizing the available natural light and is a key strategy for energy savings in commercial spaces [3]. As lighting represents a significant portion of the electrical consumption in buildings and is naturally the most common form of load, reduced energy consumption also means less load on the grid and less negative impact on the environment [2].

There are several different approaches in dealing with the issues associated with the high rise in energy demand and environmental impact. Demand-design management (DSM) is referred to a group of actions and strategies to address these challenges [4]. A few of the effective activities regarding DSM consists of; 1) shifting of non-critical usage of electricity from peak periods to off-peak periods, 2) promotion of energy efficient awareness and 3) the installation of energy management devices [5]. Behaviour-based approaches such as 1) and 2) can be achieved without capital investment. However, these strategies alone are not sufficient to meet the needed energy savings, hence the requirements for installation of energy devices [6].

The possibility of having access to natural light in indoor environments plays an important role in the effort to reduce the energy consumption. Studies have also showed a correlation between the physiological and psychological well being and exposure to daylight [1]. But natural lightning may also cause discomfort among the occupants, thus its important to evaluate an objective standard for visual comfort to achieve satisfaction among occupants [1]. A well-designed and appropriate lighting system increases the working performance by providing the requirements for visual comfort [9].

There are several strategies to achieve high energy savings for lighting systems, such as energy efficient lamps, appropriate illuminance level design, occupancy-based control and dimming control [7]. For dimming control strategies, placement of light sensors are crucial to achieve high energy savings. Improper positioning of sensors lead to poor performance of the system in achieving an energy efficient building [7]. Although lighting designers discourage the positioning of light sensors inside skylight or in direct exposure to light from fixtures, they do not imply a generalized approach how the optimal positions for light sensors can be determined [3].

The difficulties that arises with dimming control strategies are mainly linked to the con- siderable effort required to properly place and calibrate the photosensor system [10]. The unreliability of such control systems prevents the widespread acceptance of this technology in the commercial sector [10]. However, research by Li et. al [11] indicated that energy savings for electric lighting exceeds 30 % with the application of high frequency dimming controls. A study conducted in Istanbul evaluated the energy savings for a

(7)

daylight responsive system in office buildings. This study demonstrated energy savings reaching 20 % in December but increased to 47 % in June and July [12]. Furthermore, the energy savings remained approximately the same during mixed days and clear days [12]. Another study showed that the energy savings in terms of electric lighting reached approximately 70 % when using dimming lighting control system for a daylit corridor [13]. Proper designing and daylight linked dimming control systems can also reduce the energy consumption caused by cooling demands during the warmer months of the year, the potential to phase down air-conditioning systems [10]. These are major reason for further research and studies within the field of lighting control strategies.

A general rule of thumb is that a photosensor sensor should be located at a distance from the window equal to approximately two-thirds the depth of the daylight-control zone [14].

However, this rule of thumb does not account for the positioning and characteristics of the artificial lighting that may have an impact on the performance of the photosensor.

The light sensors should not have a direct view of the electric lightning they control, and be shielded from direct light from the windows [15]. Twumasi et. al [5] proposes a sensor positioning method based on the fundamental principles of illumination. The limitations with this method are mainly linked to the neglection of reflective lighting and luminaire characteristics. These issues may best be solved with the adoption of lighting simulation tools. The ideal location for a light sensor would be on the working plane to monitor the light level on the relevant task area. But such location is not practical as it would be disturbed or shaded by occupant activities in the room, hence the placement on the ceiling. Doulos et. al [16] emphasis that the suitable positioning of a light sensor may be defined in regards to the correlation of the lighting levels between ceiling and working plane. As long as the correlation remains relatively constant for a given position, the position may be considered proper. This thesis will evaluate the optimal light sensor positioning based on multiple criteria analysis with the aid of a lighting simulation software. Additionally, a lighting design method will be applied with the aim to improve the overall visual comfort.

(8)

1.1 Purpose and goals

The purpose of this study is to investigate a suitable sensor positioning method based on illuminance-based metrics in a classroom. In addition, the current lighting arrangement in the classroom will be compared to test designs based on a lighting design method with the aim to improve visual comfort. Lastly, the potential energy savings over the lifespan of the lamps will be estimated based on the decline in lamp performance and daylight contribution over a year. The most important questions at issue can be divided into three parts.

• Evaluate the current lighting design and visual comfort parameters and examine a method for improved designing.

• Perform a sensor positioning method and test the credibility by using a lighting simulation software.

• Calculate the potential energy savings due to daylight harvesting and determine the optimal light sensor positions in consideration to the natural change in daylight over a year.

1.2 Thesis outline

This thesis will begin by evaluating the term visual comfort and the photometric terms relevant to the issue. The two most common artificial light sources in commercial use will be analysed based on photometric and electric properties, with more focus towards the fluorescent lighting technique. The thesis include literature review on different lighting control strategies and the compatibility with artificial lighting. The European standards regarding visual comfort for classrooms will be presented and applied as the threshold values for the lighting simulations. A method to improve the lighting arrangement will be put to test with the aid of a software simulation program. The simulation model will be compared to field measurements for veracity analysis. A proposed method for sensor positioning will be evaluated and its credibility will be analysed based on the reference simulation values. Lastly, the potential energy savings due to daylight harvesting will be estimated and the optimal light sensor positions will be determined in consideration to the natural change in daylight over a year. The light sensor positions obtained by the daylight simulations will be compared to the positions attained by the sensor positioning method, hence examining the credibility of the method.

(9)

2 Theory

The following chapter will review the technical aspects in the field of lighting, as well as the influence on health, performance and satisfaction of occupants. The term visual comfort will be evaluated and a method for optimal lighting arrangement will be reviewed.

A sensor positioning method regarding lighting control strategies will be analysed for optimal energy savings and potential applications. Lastly, recommendations and European standards within the field of lighting will be presented.

2.1 Light

The term light has many different meanings depending on the context and application.

According to Zwinkles [17], light can be broadly defined as electromagnetic radiation with wavelengths approximately between 10 nanometer and 1 millimeter. This is known as optical radiation which describes the ultraviolet, visible and infrared regions of the electromagnetic spectrum. The International Lighting Vocabulary (ILV) [17] describes a light stimulus with the following two definitions:

• It is a characteristic of all sensations and perceptions that is specific to vision.

• It is radiation that is considered from the point of view of its ability to excite the human visual system, i.e., it is limited to electromagnetic radiation with wavelengths in the visible spectral region between approximately 380 and 780 nm.

Zwinkles [17] furthers emphasizes that the term light is not synonymous with electromagnetic radiation, as electromagnetic radiation also involves X-ray, gamma-ray and the radio range that with wavelengths outside the interval (10 nm - 1 mm). Electromag- netic radiation is defined as emission or transfer of energy in the form of photons and electromagnetic waves. Photon is a quantum in electromagnetic radiation and is usually associated with radiation that is characterized by one wavelength or frequency. Elec- tromagnetic wave is a transverse oscillation of inextricably linked electric and magnetic fields traveling through space. Figure 1 below illustrates the electromagnetic spectrum which is a graphical representation of electromagnetic waves arranged according to their wavelength.

(10)

Figure 1 – The regions of electromagnetic spectrum which highlights the optical spectrum.

[17]

The term light will from now on be refereed to light within the visible spectrum between 380 nm to 780 nm.

2.2 Light and health

According to P.Šujanová et. al [1], many studies claims that light has a direct impact on health factors. This involve both the physiological and psychological well being of the occupants, which has been proven in research during the last decades [1]. Insufficient or inappropriate light can lead to distortions of internal biological rhythms. This may affect the performance, comfort, safety and health of the occupant. The exposure to adequate lighting promotes the synchronization of the human circadian rhythms which is linked to hormone secretion. These are major motivational factors for further research within the field of lighting [1].

2.3 Visual comfort

The European standards EN 12665, defines visual comfort as ”the subjective condition of visual well-being induced by the visual environment" [18]. P.Šujanová et. al [1] emphasis that this demands a well-designed lighting system that provides sufficient illumination to enable movement and ensure safety. Parameters used to monitor visual comfort can be divided into the quantitative physical measures of the environment and qualitative aspects of vision [1]. The quantitative physical measurements are refereed to luminance, illuminance, glare and daylight provision. The qualitative aspects are refereed to color rendering, uniformity, distribution and the spectral composition of radiation. Its important that a significant part of the illumination is being provided by daylight openings, which gives the possibility for an outdoor view and also contributes to the psychological well-being [1].

(11)

2.4 Photometric terms and units

This section reviews relevant photometric terms in standard lighting calculations. Yong Xu [19] mentions that in the International System of Units (SI), there are a number of photometric quantities. Among these are luminous flux, luminous intensity, illuminance, luminance. Qualitative aspects such as illuminance uniformity, color temperature and color rendering index will also be evaluated as they are important parameters to monitor visual comfort. Table 1 below explains the relationship between the units for luminous flux, luminous intensity, illuminance and luminance.

Table 1 – Photometric quantities and respective symbols and units. [20]

Quantity Symbol Unit

Luminous flux φ_v lm = cd ·sr Luminous intensity Iv cd = lm/sr

Illuminance Ev lx = lm/m²

Luminance Lv cd/m²= lm/(m²· sr)

Whilst quantified properties such as luminous flux and luminous intensity are directly connected to the specific light source, illuminance and luminance are associated with the provision of light received by an surface or passing through an object [19].

2.4.1 Luminous Flux

Choudhury [21] describes luminous flux as the measure of the perceived power of light.

The SI-unit of luminous flux is lumen (lm) which is defined as the luminous flux of light produced by a light source that emits 1 cd of luminous intensity over a solid angle of 1 steradian (sr).

2.4.2 Luminous Intensity

Luminous intensity is the basic photometric value expressing the capacity of a point light source to provide illumination in a given direction [22]. The SI-unit of luminous intensity is candela (cd) which is defined as the quotient of the elementary luminous flux by the elementary solid angle in which it is propagated.

2.4.3 Luminance

According to Alrubaih et. al [10] luminance is defined as ”a measure of the luminous flux that passes through or is emitted from unit area of a surface". The SI unit of the luminance is candela per square meter (cd/m2) and indicates the brightness of a surface.

The luminance is therefore the ratio of the luminous intensity and the projected area of the light source. It can be described as the formula eq. 1 below [20].

(12)

L_v = I_v

A_s· Cosθ, (1)

where Lv is the luminance at the projected area As and the θ is the angle between the directions of the incident light and the normal to the surface.

2.4.4 Illuminance

Illuminance can be defined as the total luminous flux incident on a surface, per unit area [19, 23]. It is analogous to the power density of the radiated light, but adjusted by the luminosity function. In SI-derived units, it is measured in lux (lx) or lumens per square meter (lm/m2). Its important to distinguish the difference between illuminance and luminance. As illuminance reefers to the amount of light falling onto a given surface, luminance refers to the light passing through an object [24]. Illuminance is the most common photometric quantity to monitor sufficient lighting in a specific area. The relationship between the illuminance at a specific point and the luminous intensity from a light source can be formulated in eq. 2 below [25].

E_v = Iv· Cosθ

d² , (2)

where Ev is the illuminance at a specific point, Iv is the luminous intensity from the light source and d is the distance between the light source and incident surface. The equation is based on the cosine law where the the illuminance on a surface is proportional to the cosine of the angle θ between the directions of the incident light and the normal to the surface. When calculating the horizontal illuminance on a working plane, the following constraints applies, 0^◦≤ θ < 90^◦.

2.4.5 Uniformity

Lighting uniformity can be defined as quality of the overall illuminance distribution. Its a ratio of the minimum illuminance and the average illuminance in a certain area [26].

Uniformed lighting allows the occupant to perceive the environment without sudden breaks of lighting level that causes discomfort. The indicator uniformity is frequently used due to the simple determination based on illuminance measurements [27]. A study conducted by Lee et. al [28] indicated that improved uniformity can be achieved by increasing the gap between the artificial lighting sources and the targeted plane.

2.4.6 Color Rendering

Color rendering index can be defined as "the effect of an illuminant on the color appearance of objects by conscious or subconscious comparison with their color appearance under a reference illuminant" [29]. This index indicates how well a light source can ren- der a color on the observed object. Natural color rendering with the highest index is the

(13)

sun, which is 100. The desired index depends on the environment and demands required in the area. Artificial lights aims to achieve index higher than 90, in office buildings for example, whilst residency and outside lighting can be slightly lower.

2.4.7 Correlated Color Temperature

The correlated color temperature (CCT) can be defined as ”a luminous source that is the color of the energy emitted, represented by the heating of a black body" [30]. Figure 2 below illustrates the heating of a black body, corresponding in different colors. Heating between 2000 kelvin and 3000 kelvin results in a red color known as "hot". 3000 kelvin to 4000 kelvin results in a more yellowish color and 4000 to 5500 kelvin results in a white and more neutral color. Above 5500 kelvin, the emission is considered blue. A study made by Ball State University, Indiana indicated that higher CCT resulted in increased alertness, attitude and energy level among students [31].

Figure 2 – Describes the typical colors and their corresponding color temperatures. [32]

2.5 Glare

Glare is defined as "the sensation produced by luminance within the visual field that is sufficiently greater than the luminance to which the eyes are adapted to cause annoyance, discomfort, or loss in visual performance and visibility [23]". The two most common type of glare issues are discomfort glare and disability glare [10]. Glare is recognized as an very important issue in providing visual comfort.

2.5.1 Discomfort glare

Alrubaih et. al [10] claims that the most common form of glare is the discomfort glare which have physiological causes that are not fully understood. However its a sensation

(14)

of annoyance caused by very high or non-uniform distributions of brightness. Discomfort glare can cause discomfort and eyestrain if the occupant is exposed in a longer period of time. Important key factors to minimize the potential of discomfort from glare are; the size, luminance, and number of glare sources and the geometry of the sources in relation to the eye and task plane.

Kruisselbrink et. al [27] states that discomfort glare is relatively difficult to identify as its a visual sensation, which cannot be measured directly. Discomfort glare is thus associated with psychological factors, unlike the disability glare which is associated with physiological factors.

2.5.2 Disability glare

The other and more acute form of glare is the disability glare which obstructs the field of vision. There are no accurate models to predict the disability glare but important factors includes relative brightness, size of the area affected and composite intensity of all the glare sources within the field of view [10].

2.5.3 Unified Glare Rating

According to Akashi et. al [33], the unified glare rating (UGR) is a glare evaluation based on generally accepted parameters influencing discomfort glare. Sorensen proposed the UGR based on the following formula (eq. 3) [33]:

U GR = 8log10·0.25 L_b

X L²· ω

p² , (3)

where Lb is the background luminance [cd/m2], L is the luminance of the luminous parts of each luminaire in the direction of the observers eye [cd/m2], ω is the solid angle of the luminous parts of each luminaire at the observers eyes [sr] and p is the position index.

The UGR for discomfort glare can be categorized in table 2 below.

Table 2 – The UGR index and respective category [33].

No. Categories Glare index

7 Just intolerable 28

6 Uncomfortable 25

5 Just uncomfortable 22

4 Unacceptable 19

3 Just acceptable 16

2 Perceptible 13

1 Just perceptible 10

0 Imperceptible 7

(15)

Furthermore, the unified glare rating is a dimensionless parameter which provides information about the degree of psychological glare of a lighting installation in an indoor space and is used in lighting calculations to prevent visual discomfort [34].

2.6 Daylight

Baker et. al [35] defines daylight as "the combination of the diffused light from the sky and sunlight", whereas Manning [36] indicated the term daylighting as any method by which natural light is brought into a room to replace or supplement artificial lighting.

The sun is the source of natural light and the access to sunlight at a particular point is determined by the path of the sun. With the use of the solar azimuth and solar altitude angles, a reference point on the surface of the earth can be set to decide the position of the sun [10]. The three common light condition to consider are overcast, partly cloudy and clear skies. More specific classification are illustrated in section 2.6.3. Daylight penetration and outside view are the two main aspects regarding daylighting in buildings [27].

2.6.1 Daylight penetration

Daylight penetration is given depending on the fixed window openings, weather, time and geographic orientation. A surplus of daylight might lead to discomfort, however it is possible to optimize the daylight penetration without causing discomfort among occupants. There are three common techniques that many modern buildings are applied with; dynamic sun shading (sun screens), brightness control (blinds) and/or smart glazing integrated in the facade [27]. Smart glazing may be defined as glass whose light transmission properties can be altered using electrical voltage. Sun shading and smart glazing are used to block the direct solar radiation, and also prevent overheating in the room and glare issues. These type of system are usually fixed, but the dynamic systems that follow the trajectory of the sun are available. Brightness control usually has a dynamic character and the blinds are easy to adjust automatically or manually, permitting optimizations. As daylight penetration can cause discomfort if not optimized correctly, by using a control algorithm it may be used more advantageous increasing visual comfort and reduce energy consumption.

2.6.2 Outside view

Kruisselbrink et. al [27] further claims that many studies suggests that a high quality outside view and visual contact with the surrounding environment can influence the performance and satisfaction with the work, and even result in general health benefits. The view outside is influenced by the size and shape of the window openings and the quality of the view, having a subjective nature. Rather than aspects of the built environment, a high quality outside view generally consists of natural aspects.The quality of view is also very dependent on the fixed location Its largely subjective but methods have been developed to assess the subjective view quality objectively. Although its an complex subject

(16)

and further validation are required for these methods. Its important to mention that the outside view can also be obstructed by sun shading, brightness control and smart glazing.

2.6.3 Sky conditions

According to Igawa et. al [37], sky condition may be classified into five categories based on the sky index (SI). Table 3 below summarizes the sky index for the different sky conditions.

Table 3 – Sky conditions and their classified category [37].

Sky condition Sky index (SI)

Clear sky 1.7 ≤ SI

Near clear sky 1.5 < SI ≤ 1.7 Intermediate sky 0.6 < SI ≤ 1.5 Near overcast sky 0.3 < SI ≤ 0.6 Overcast sky SI ≤ 0.3

Hayashi et. al [38] claims that the intermediate sky may be defined as the luminance distribution which represents all sky condition except the clear sky and overcast sky conditions in daylight calculations. Nakamura et. al [39] states that the intermediate sky is a kind of the average sky, developed from a long period of measurements, describing average weather conditions.

2.7 Artificial light

In contrast to daylight, artificial light is human-made and can emerge from sources such as fire, gaslight, candlelight and electric lamps which is the most common source of artificial light in today’s society [40]. In this study, artificial lighting will be refereed to the modern electric lamps. The five most common types of artificial light sources are incandescent lamps, compact fluorescent lamp, fluorescent tube, discharge lamps and light-emitting diode (LED) [41]. The fluorescent tube, alongside LED, are the standard lighting solution within school environment [31]. Although the LED lamp is the most energy efficient technology among all kind of illumination, it remains relatively expensive compared to the fluorescent tube [42]. The fluorescent tube technology is still considered energy efficient and provides a long lifetime of the tubes, thus an attractive candidate within the commercial sector [43].

Energy efficiency can be defined as the consumption of less energy in the process but guarantee the same energy service [30]. In the field of lighting, the energy service cor- responds to sufficient illumination in the area. Thus, an important factor to consider when choosing a suitable type of lamp is the luminous efficacy. According to Choudhury [21], this is a measure how well a light source produces visible light in terms of electric efficiency, and is defined as the ratio of luminous flux to power (see eq. 4 below).

(17)

K = φ_v

P , (4)

where K is the luminous efficacy, φv is the luminous flux produced by the light source and P is the operating power.

Furthermore, the energy consumption can be described with the fundamental equation below [2].

E = P · t, (5)

where E (Wh) is the consumed energy, P (W) is the installed effect and t (h) is operating time of the artificial lighting. Thus, the energy consumption can be reduced by either lowering the installed effect, reduce the amount of operating hours or both. Reducing the amount of operating hours can be done efficiently with the application of a scheduling time switch system or occupancy sensors, whilst the power may be reduced by utilising the available daylight in the room, thus integrating a daylight-linked system which regulates the power output and light level [2]. These strategies will be discussed more in chapter 2.8.

2.7.1 Fluorescent lighting and Ballasts

Fluorescent lamps are defined as lamps which emits fluorescent light, usually generated by irradiation of a phosphor with light from an electric gas discharge [44]. According to Energy gov [45], there are two general types of fluorescent lamps, the compact fluorescent lamp (CFL) and fluorescent tubes. The CFL is commonly found with integral ballasts and screw bases and these are popular lamps often used in household fixtures. The fluorescent tube lamps are typically used for task activities and for lighting large areas in commercial buildings.

In fluorescent lighting systems, a ballast regulates the current to the lamps and provides sufficient voltage to start the lamps [46]. Without the ballast, a fluorescent lamp connected to high voltage power source would rapidly and uncontrollably increase its current draw. The ballast must supply high voltage to establish an arc between the two lamp electrodes during the start. The ballast quickly reduces the voltage and regulates the electric current to produce a steady light output once the arc is established [46].

Larger fluorescent lamps usually use an inductor coil as the ballast which makes it possible to operate at the usual AC current [44]. The inductor also provides high voltage spike at start up. Instead of a simple inductor, a electronic circuit can be used as a ballast and for start up.

The two most common types of ballasts are magnetic ballasts and electronic ballasts.

The electronic ballasts consists of electronic components that operate lamps at 20-60 kHz, compared to magnetic ballasts that operate lamps at linear frequency of 60 hz

(18)

[47]. Fluorescent T5 model requires a electronic ballasts which will therefore be more interesting in this study (see appendix A.1). There are several advantages with a electronic ballast such as; more compact device, possibility to operate at DC and it allows higher power efficiency which essentially eliminates flicker [44]. Other advantages such as 10-15 % increased lamp efficacy, quieter and half the power loss factor compared to the magnetic ballast further favours the electronic ballast [47]. The flicker can be described as alternating current that modulates the light output and becomes perceptible.

Fluorescent lamps operating with magnetic ballasts has a flicker index ranged between 0.04-0.07. Most electronic ballasts has a flicker index below 0.01 due to its high frequency operation. High index values indicates higher probability for flicker. For these reasons mentioned, high efficient electronic ballasts are promoted on the market for the significant energy savings over magnetic ballasts, in addition the advantages favouring visual comfort [47].

Transformers and inductors are essential components of the electronic ballasts. Fluores- cent lamps can be dimmed by adjusting the switching frequency, the magnitude of the source voltage or value of the inductor [48]. However, variation of the switching frequency disrupts the performance of the lamp when operating at low power. Instead of shifting the switching frequency, the control current (Idc) are modulated to met the desired lamp power. This improves the dimming characteristics and eliminates burst dimming [48].

Studies conducted by Kontaxis et. al [49] evaluated the consumed power relative to the light output for four different T5 luminaries. Their study found that the relationship between the consumed power and light output followed a similar linear pattern for all T5 luminaries.

2.7.2 Luminaries and polar curves

According to IES Lighting Handbook [50], luminaries can be defined as "a complete lighting unit consisting of a lamp or lamps together with the parts designed to distribute the light, to position and protect the lamps, and to connect the lamps to the power supply". This means that the term luminaire includes the lamp and all the components directly affiliated with the lumen distribution, protection and positioning of the light unit. Luminaries consists of shielding medias which affects the characteristics of the light, for example; translucent diffuser, parabolic louver or reflectors [51]. The purpose of these shielding medias is to minimise glare occurrence. Many luminaries are designed with light directed upwards, thus resulting indirect lighting and less issues related to glare.

Polar curves describes the light distribution of a luminaire. The polar curves defines the variation of luminous intensity with angle of emission in a 3D C 0-180 and C 90-270 plane from the centre of the luminaire [52]. Furthermore, the polar curve differentiate depending on the shape of the luminaire fixture and type of lamp being used.

(19)

2.8 Lighting control strategies

Research shows significant energy savings with the applications of various types control schemes [2]. The classic manual lighting controls depends mostly on occupant behaviour, occupant pattern and general awareness about energy saving [53]. Thus, researchers has been focusing on developing efficient automatic control systems to increase energy savings whilst maintaining visual comfort. Automatic schemes vary in complexity and technology. On a basic level, the automatic controls can be used to switch on or off the lights which is referred as daylight-linked switching [54]. But more precise techniques have the capacity to control the level of illumination based on requirements through dimming, which in turn requires a dimmable ballast [54, 2]. The latter technique is referred to as daylight-linked dimming [2]. Daylight-linked systems can also be divided based on algorithm of control, these are open loop system and closed loop system which will be discussed below.

Another strategy widely used is the occupancy-based control scheme which in turn can be integrated with daylighting control schemes for optimal energy savings [2]. All different strategies will be reviewed below.

2.8.1 Key factors for daylight-linked systems

The implementation of daylight-linked control systems requires a comprehensive evaluation of key factors that affects the performance of the system, as it may turn costly and provide low energy savings in a poorly commissioned light system [2]. The general key factors that affect the daylight availability are listed below:

• Sky conditions in the current location.

• Window properties, such as positioning, orientation and window glazing.

• Surrounding obstructions, such as building, trees or other structures that may obstruct daylight entering the room.

In addition to the general key factors, Mohammad et. al [2] emphasis the importance of proper tuning of control parameters. This include suitable lux levels, delay settings and placement of sensors. Suitable lux levels and delay settings refers to the control schemes to maintain a pre-set level of light intensity from the lamps based on the available daylight in the control zone. Mohammad et. al [2] further claims that the placement of sensors is important in order to gather accurate information regarding the light distribution. Theoretically, the ideal positions to place sensors would be the task area but such placement is not practical. Thus examining methods for optimal positioning of light sensors is crucial for optimal performance of the control system. The positioning of several inexpensive sensors can provide improved accuracy in detection, which in turn allows shorter time delay settings, thus further energy savings. Sensors are not recommended to be placed in positions exposed to direct sunlight or external glare [2]. Thus, its not suggested to place them nearby windows as it may give misleading values which does not represent the illumination at task area [55]. Two types of daylighting control

(20)

schemes utilized to reduce the electric energy consumption are dimming control systems and switching control systems [23].

2.8.2 Daylight-linked dimming control

Continues dimming control usually involves a sensor positioned at the station point best suited [10], or multiply sensors spread out depending on the amount of working spaces [7]. The sensors responds to sunlight or natural light in order to maintain the set indoor illuminance level [10]. If the exterior illuminance exceeds the set indoor illuminance level, the artificial lights are dimmed, and vice versa. The photosensors adjust the electric lighting level depending on the amount of light received [10]. Important factors influencing the performance of these systems are the sensor placement, commissioning and hardware quality [56]. This type of control system is best suited for daylight applications as it increases the energy savings, extends the life span of the lamps compared to switching control systems and improves the comforts of the occupants [57]. The cons with this type of system are mainly the expenses as they require special lamps, ballasts and more advanced controls [10]. Colaco et. al [52] further emphasis that there is not much deviation in the light distribution contour shape at different dimming levels.

2.8.3 Daylight-linked switching control

According to Topalis et. al [56], simple switch on/off technique switches the electric light off when the daylight illuminance in the station point is reached, and turned on again when the daylight illuminance drops below the control value. The main problem with photoelectric switches is the users reaction to its operation [56]. Another problem with switching controls is that the daylight levels are fluctuating and the lamp life reduces by constantly switching it on and off, it may also cause frustration among occupants [56].

2.8.4 Comparison between dimming and switching controls

According to Mohammad et. al [2], there are several factors determining which technique is more suitable. As the cost of installation is one of the key factors in adopting technologies, switching control is generally cheaper easy to install. However dimming control represents greater accuracy in control and may seen as less obtrusive to occupants due to the gradual change between light levels. Its also important to evaluate the daylight conditions when choosing technique. Switching control may be suitable with consistent daylight contribution throughout the day whilst dimming control may be more suitable with higher variation of daylight throughout the day [2]. Additionally, the occupant activity must be evaluated as switching controls are recommended for non critical tasks in circulation areas or hallways whilst dimming controls might be preferable in classrooms or offices [2].

(21)

2.8.5 Open loop or closed loop

This entire section is referred to Mohammad et. al [2]. Daylight-linked systems can be divided based on the algorithm of control. These are open-loop and closed-loop systems.

A closed-loop system continuously detects lux levels of the control zone, which includes lighting from both the artificial lights and daylight. The change in the light levels of the lamps due to the availability of daylight is fed-back to the control system continuously, and it can make necessary adjustments based the feedback. Figure 3 below illustrates the closed-loop system.

Figure 3 – Daylight-linked closed loop algorithm [2].

On the other hand, the open-loop systems does not receive any feedback from the level of artificial lighting, it only detects available daylight levels. Based on the level of available daylight, it sends corresponding signal to the controller to provide corresponding lamp output. Figure 4 below illustrates the open-loop system.

Figure 4 – Daylight-linked open loop algorithm [2].

Choosing open or close loop system and placing the sensors accordingly is essential for successful commissioning. When the goal is to control multiple control zones with a single sensor, open-loop system is more preferable. Open plan office spaces are good targets for open-loop daylight system implementation. When the sensor is used to control a single control zone or comparatively small areas, like private offices, closed loop systems can be very effective [2]. As students perform stationary and critical tasks, the artificial lighting system will require to maintain a set light level target, thus a closed loop system may be more appropriate in classrooms [14].

(22)

2.8.6 Occupancy-based control

According to Mohammad et. al [2], occupancy-based sensors are sensors that senses the motion of the occupants. In theory, the manual switching controller contributes to a higher energy reduction compared to the occupancy-based sensors in office buildings.

The main issue with manual switch is the attitude of the occupants to leave the lights on when leaving the room. Hence automatic lighting control system such as occupancy-based sensors are a great way to mitigate the wasted energy for lighting in buildings.

2.8.7 Mixed strategies

Mohammad et. al [2] further states that mixed strategies which implements different technologies may have the best possible outcome. As shown in the previous sections, each control technique has their unique characteristics that may all be advantageous depending on the environment and purpose of use. Yet these technologies often fail to provide satisfactory performance due to the shortcomings associated with the particular technology. Mixed strategies means a combination of different technologies suitable for the situation in order to maximize the energy-savings without compromising the satisfaction of the occupants. Studies shows that a combination of technologies gives substantial improvements in performance in terms of accuracy and energy-savings.

2.9 Lumen method

The lumen method is a common approach to compute illuminance-based metrics for lighting designs [7], recommended by Ref. [50]. The following paragraph and listing below are provided by Dialux knowledge base [58]. The lumen method serves to provide approximate calculations when planning indoor lighting systems. Its used to determine the number of luminaries and lamps required to achieve a desired level of illuminance. The result generated by the lumen method depends on a number of factors and assumptions which may lead to results deviating from reality. The more deviating the reality is from the assumptions, the less accurate calculations. The basic assumptions underlying the lumen method are the following:

• Rectangular room

• Ratio of length to width = 1.6:1, with a maximum of 4:1

• Completely empty room

• Uniform reflectance and completely diffuse reflection properties of the perimeter surfaces

• Regular luminarie configuration throughout the room

• Uniform distribution of luminous flux over all areas

• In the case of fluorescent lamps, luminaire axis = room axis

(23)

The first step in the lumen method is to determine the utilisation factor from the table of photometric data [59]. The utilisation factor is the ratio between the total lumen emitted by the light source and received on the working plane and can be determined by establishing reflectance values for ceiling, walls and the floor and calculating the room index with the following formula (eq. 6):

k = a · b

h_k· (a + b), (6)

where k is the room index, a is the length of the room in [m], b is the width of the room in [m] and hk is the useful height between the luminaire and working plane. Alzuhairi [60] states that the reflectance values for a standard room may be set to C = 0.7, W = 0.5 and F = 0.2, where C is the ceiling, W is walls and F is the floor.

With given reflectance values for ceiling, walls, floor and determined room index, the utilisation factor ηb can be found in the table of photometric data supplied with the luminaries [59]. The utilization factor for the specific lamp examined in this study could not be found. However, values could be obtained for similar fluorescent lamps in order to provide adequate accuracy (see appendix A.2).

The maintenance factor Mf depends on several factors determining the performance of the luminaire. It can be formulated in eq. 7 below [59]:

M_f = LLM F · LSF · LM F · RSM F, (7)

where LLMF is the lamp lumen maintenance factor - the reduction of in lumen output after a specific operating hours. LSF is the lamp survival factor - the percentage of lamp failures after specific burning hours. LMF is the luminaire maintenance factor - the reduction in light output due to dirt deposited on or inside the luminaire. RSMF is the room maintenance factor which describes the reduction in reflectance due to dirt deposition on the room surfaces. Wagiman et. al [7] claims that the Mf may be set to 0.8 in the commercial sector and that the ballast factor Bf is always considered as 1, thus often neglected in the equation.

The last step is to determine the amount of luminaries required to achieve a desired level of average illuminance in the room. This is procedure can be done with the eq. 8 below [59].

N = E · A

ηb· n · M_f · φ_v, (8)

where N is the amount of luminaries which will be rounded up to obtain a whole number of luminaries for a logical configuration, Mf is the maintenance factor, E is the desired

(24)

average illuminance, A is the area of the room, n is the number of lamps per luminaire, φv is the rated luminous flux of a lamp and ηb is the utilisation factor.

This equation can be transformed to estimate the average illuminance based on the number of luminaries.

E = η_b· n · φ_v· N · M_f

A , (9)

In terms of positioning of the luminaries, Alzuhairi [60] states that in order to achieve uniform light distribution across the room, the distance between two adjacent luminaries are determined with the following formula (eq. 10).

dl = a Nr

, (10)

where dlis the distance between two adjacent luminaries, a is the length of the room and N_r is the number of luminaries in a single row.

The distance between the luminarie and its adjacent wall can be estimated by using the following formula:

dw = x · h, (11)

where dw is the distance between the luminaire and its adjacent wall, h is the height of the room and x is a factor with the following constraints:

x = 1/2 if a/b < 1.6, else x = 1/3, where a and b is the length respective width of the room.

2.10 Point-by-point method

As mentioned in section 2.8.1, factors such as positioning and the number of sensors affects the performance of monitoring accurate illuminance distribution in a room. Twu- masi et. al [5] proposes a method that determines the optimal positioning of light sensors based on the fundamental principles of illumination. This method will be refereed to as point-by-point with the purpose to determine the level of workplane lighting in an internal environment. This method is also applied by Lee et. al to calculate the illuminance distribution in an artificial plant factory [28]. Their study emphasis that large deviations may occur as the point-by-point method does not account for reflective light, hence the necessity to perform simulations or field measurements. However, Doulos et. al [16]

emphasis that the ratio of light levels between working plane and ceiling should stay relatively constant for a photosensor to operate properly. Thus the suitable positioning of a photosensor requires a linear relationship of light levels between work plane and ceiling for optimal performance.

(25)

The entire section below covering the methodology of Point-by-point is referred to the article by Twumasi et. al [5].

This solution to determine the optimal placement of light sensors is fairly easy and hence, a time efficient method. Its based on dividing the room into a number of zones or clusters, whereas each zone involves a certain number of calculation points. The amount of calculation points may be modulated depending on the size of the area. Increasing the amount of calculation points results in greater iterations and more precise results, the downside is the more time consuming process. The amount of zones determined are dependent on the number of desired light sensors to be installed. According to Wagiman et. al [7], the number of light sensors deployed should not exceed the amount of luminaries as the central controller would be conflicting, thus the numbers sensors should be equal or less than the number of luminaries in the room. To simplify, the number of zones, sensors and luminaries are all equal in this case. The average illuminance for each zone is determined and the corresponding coordinate point is the optimal sensor location.

The light contribution is assumed to only be provided by the interior artificial lighting system, thus exterior artificial light sources and daylight are neglected in the iterations.

This method can be described with the following steps:

The calculation points are determined by obtaining a set of length A = [1,2,3,...,n] and the width B = [1,2,3,...,m]. The interval length of A and width B can be adjusted depending on the amount of calculation points determined.

The set of coordinate points can be placed in matrice by obtaining the Cartesian products of sets A and B, thus G is given as

G = A × B, (12)

where G is the two dimensional matrice with coordinate points Gx, Gy evenly distributed.

The amount of luminaries denoted as L are placed in the coordinate system to its corresponding position Lx, Ly. The horizontal distance, denoted as DGxy Lbetween luminaire Lxy and calculation points Gxy are the following:

D_{Gxy L}= q

(L_x− G_x)²+ (L_y− G_y)². (13) With a given useful height hk, the total distance, denoted as dGxy L, between the luminaire and calculation points can be determined using the equation below:

dGxy L= q

h²_k+ DGxy2

L, (14)

thus, cos φGxy L could be obtained:

(26)

cosφ_{Gxy L}= h_k dGxy L

, (15)

Figure 5 below gives a more clear understanding regarding the trigonometric relationship used in this method.

Figure 5 – Trigonometric relationship used in the laws of illumination, based on eq (2) from section 2.4.4.

Iv designates the luminous intensity, L designates the luminaire, hk designates the useful height, Lx and Ly designates the position of the luminaire, Gx and Gy designates the calculation point, DGxy L designates the horizontal distance between the luminaire and calculation point, dGxy L designates the total distance between the luminaire and calculation point and φGxy L designates the corresponding angle.

The illuminance for each coordinate point can be determined by applying equation (2) from section 2.4.4 and perform iterations for each point. The luminous intensity for the luminaire is decided beforehand and can be adjusted depending on the characteristics of the light source. The total illuminance for each calculation point contributed by each luminaire is determined using the following formula:

EGxy =

B

X

L=1

EGxy L (16)

The coordinate points are grouped in to clusters K which is equal to the amount of sensors S. In order to find the most suitable position for the light sensor, the average illuminance for each cluster is determined and the closest corresponding coordinate points are the optimal location. The equations below describes the final steps of the procedures:

(27)

E_{av g},_k= 1 N

N

X

n=1

E_Gxy,_n (17)

The optimum position of the light sensor for kth cluster (Pop,k) are based on considering the minimum absolute deviation of illuminance and can be expressed as following:

P_op,k = min | ∆E | (18)

∆E = Eav g ,k− E_{k ,(x},y ), (19) where Ek ,(x,y ) is the illuminance at the xy coordinate plane.

Although this is an fairly simple and time efficient technique, the theoretical principles of this methodology is not fully accurate in practical situations, as the luminous intensity is not uniform due to the shape of the lighting fixtures and lamps. Therefore the specific luminaire polar data needs to be integrated in the iteration of cos angles as the light intensity may vary asymmetrical. If the data is not available, a curve fitting approach is suitable for the best possible approximation.

2.11 Standards and recommendations

The information in the following section is provided by BSI Standards Publication[61].

The European standard EN 12464-1:2011 specifies lighting requirements for human indoor work places which meet the needs for visual comfort. It was approved 14 April 2011 by CEN members that are bound to comply with the CEN/CENELEC Internal regulations which stipulate the conditions for giving this European standard the status of a national standard without any alteration.

The lighting design criteria for a luminous environment are based on three basic human needs:

• Visual comfort, where the workers have a feeling of well-being which also contributes to higher quality of work and productivity level.

• Visual performance, where the workers are able to perform visual tasks, even during difficult circumstances and longer periods of time.

• Safety.

The main parameters determining the luminous environment with respect to daylight and artificial light are; illuminance, luminance distribution, direction of light, light levels and color of light, colour rendering and colour appearance of light, glare and flicker.

For educational premises, the following standards applies:

(28)

Table 4 – Standards for different types of areas/activities. The average illuminance on working plane is denoted as Em, the unified glare ratio is denoted as UGR, the uniformity is denoted as U_o and the color rendering index is denoted as R_a [61]

Type of area/activity Em UGR Uo Ra

Classroom 300 19 0.60 80

Classroom for evening classes 500 19 0.60 80

Lecture halls 500 19 0.60 80

Additionally, specific requirements recommends controllable lighting for all types of areas/activities above.

Table 5 describes the relationship between the illuminance on task area and immediate surrounding.

Table 5 – Relationship between illuminance level between task area and surroundings.

[61]

Illuminance on the task area [E_task] Illuminance on immediate surrounding [E_{sur r}]

≥ 750 500

500 300

300 200

200 150

150 Etask

100 E_task

≤ 50 E_task

The background illuminance requires a maintained illuminance level 1/3 of the value of the immediate surrounding area. The immediate surrounding area requires the uniformity Uo ≥0.40 and the background area requires Uo≥0.10. Figure 6 below gives a more clear understanding regarding the minimum dimensions between the specific areas.

(29)

Figure 6 – The minimum dimensions between the specific areas

Nr. 1 represents the task area. Nr.2 represents the immediate surrounding area with a width of at least 0.5 m around the task area. Nr.3 represents the background area with a width at least 3 m adjacent to the immediate surrounding area within limits of the space.

For a typical classroom used during daytime, table 6 below summarizes the specific minimum requirements in terms of visual comfort, performance and safety.

Table 6 – Illustrates the minimum requirements for a classroom used during daytime.

[61]

Parameter Value

Etask 300

Utask 0.60

E_{sur r} 200

U_{sur r} 0.40

E_back 66.7

U_back 0.10

UGR ≤ 19

Ra 80

(30)

3 Method

The following chapter involves the methods regarding the case study. The classroom N380 in the Faculty of Science and Technology at Umeå university has been evaluated based on lighting data from the current sensors, drawings and dimensions of the room, the current lighting and controls systems being used. The methodology can be divided into three major parts. The first part involves examining a more optimal artificial lighting design to improve the visual comfort in the classroom. This was done with the aid of the simulation software program Dialux Evo. Dialux evo will from now on be designated as only Dialux.

The second part involves performing a sensor positioning method denoted as "Point-by- point" to determine the optimal placement of light sensors in regards to illuminance based metrics and light distribution. The credibility and accuracy of this method was analysed based on the reference values obtained by Dialux simulations. The final part includes integration of daylight data from the software program Dialux in order to determine the optimal positioning of light sensors in regards to the change in daylight activity. Thus, also estimate the potential energy savings by dimming the artificial lighting. Figure 7 below illustrates a schematic overview regarding the methodology.

Figure 7 – A schematic overview consisting of the major stages in this study.

(31)

3.1 Classroom N380

The classroom N380 has a capacity of 40 persons and its mostly used during daytime. Its equipped with 8 luminaries whereas each luminaire contains three lamps of type Philips, Master TL5 HE, 28W/830. More information regarding the technical specifications about the lamp can be found in table 13, appendix A.1. The luminaries contains reflectors and diffusers for the light directed downwards to scatter the light and avoid glare issues. Two lamps for each luminaire are placed above the reflector with light directed towards the ceiling. By observations, four lamps directed towards the ceiling for two luminaries were out of function. Furthermore the lighting is controlled by a mix of manual switch and occupancy based detector. The room consists of seven windows oriented to the west, with an additional seven smaller windows. Furthermore, no larger objects or buildings nearby can be considered to have significant impact on the daylight availability.

Figure 8 – The classroom N380, Umeå University.

(32)

3.2 Construction and drawings

The first step in evaluating the lighting system was to construct a model in the software program Dialux. This was done by importing the CAD-file provided by Akademiska hus, then define the constant input parameters for dimensioning.

3.2.1 CAD drawings and dimensions

The CAD-files were necessary to obtain the correct dimensions of the rooms, windows and exact placement of the current luminaries. However, real measurements was performed to ensure that the CAD-files correspond to reality as changes might have taken place without file updates. The height of the ceiling was also measured, as well as window dimensions. Figure 9 below shows the imported CAD-plan.

Figure 9 – CAD-plan for the classroom N380.

The designation LP3 represents the luminaries. The file matched well with reality, except for one luminaire that has been positioned closer to the door. The dimensions of the room, windows, tables and luminaries are can be found in table 14, appendix A.3.

3.3 Capabilities of existing software tools

The relevant software programs used in this study are Dialux, Matlab and Microsoft Excel. They all serve meaningful purposes in terms of simulations, iterations and illustrations of result.

Dialux is a developed simulation software that is widely used to design, calculate and visualize light for rooms, entire buildings and outdoor areas. Dialux offers a multi- tude of features and functions including integration of daylight data in different sky

(33)

conditions[62]. The simulation software has been used in various studies in terms of lighting design and illuminance evaluation [7, 8, 9, 63, 64, 65, 66].

Matlab is a software tool widely used by engineers and scientist provided by Mathworks [67]. Matlab combines a desktop environment tuned for iterative analysis and design processes with a programming language that expresses matrix and array mathematics directly. Matlab allows smooth computations, numerical iterations and plot illustrations for various problems. This software tool has been used in this study to perform numerical iterations for the sensor positioning method and establishment of functions for energy savings.

Excel is a program provided by Microsoft with functions such as iterative calculations and simple data management tasks [68]. The program is a practical tool which provides clarity in large set amount of data, including graphical application suitable to illustrate results in diagram formats. Microsoft Excel has mainly been used in this study for documentation and management of data provided by Matlab and Dialux.

3.4 Assumptions and limitations

This section will present the major assumptions and potential limitations with this study.

The assumptions and limitation associated with the daylight simulations will be discussed separately in section 3.8.

As Dialux has been used in previous scientific studies, the simulation software has been assumed to provide reasonable and trustworthy data corresponding well with reality. In contrast, as only Dialux has been used in this study, it may be considered a limitation as the use of additional lighting simulations tools might have confirmed/strengthen or question the results obtained by the Dialux simulations. The reflection factors adjusted to the correct values for a standard classroom in Dialux might deviate from the reality of classroom N380, thus this assumptions might have influence on the results obtained by the lighting simulations in relation to the real case scenario. Furthermore, the manufacturers tab in Dialux had limited amount of luminaries to import. The luminaire model imported was based on achieving similar characteristics as the luminaries in N380. However, minor differences of the models might have impact on the lighting distribution.

The functions for the dimming potential in section 3.5.3 are based on three major assumptions. 1) The use of dimmable ballasts compatible with the lighting system, which is crucial for a daylight-linked dimming control system to even operate. 2) A preheated start of the ballast (see figure 33 in appendix A.1), which results in extended lifetime of the fluorescent lighting system due to the reduced decline in installed luminous flux.

3) The method to determine the potential energy savings in 3.8.1 are based on the assumption of a closed loop algorithm as this type of control system is recommended when controlling a single zone like a classroom. The major limitation with closed loop algorithm is the unexpected behaviour (oscillating and switching) that can occur if not set up correctly. The classroom was assumed to be used 8 hours every day with the exception of weekend and the summer months. This may cause limitations associated with the end

(34)

result as the classroom will most likely not be used monotonous throughout the year.

For example, an increased amount of operating hours in January in relation to a reduced amount of operating hours in May would result in decreased energy savings due to less daylight contribution despite the same amount of operating hours annually. Furthermore, the impact of the occupancy detector has been neglected in the determination of energy savings. This aspect would require a more comprehensive analysis on student behaviour and the use of the classroom. The classroom might also be used outside of scheduled time which adds complexity to the issue as no data regarding the current amount of operating hours of the lighting system was available.

Table 7 below summarizes these major assumptions and potential limitations associated with this study.

Table 7 – Assumptions and potential limitations with this study. The limitations are strongly associated with deviation from the reality.

Assumption Limitation

Simulation software Trustworthy data Only used one program

Reflection factors Theoretical standard classroom Potential deviation from real case Luminaire model Similar characteristics Minor differences from real case Ballast Dimmable and preheated start Usually more expensive

Control algorithm Closed-loop More complex system

Operating hours Monotonous throughout the simulations Deviate from reality Occupancy detector Neglected in calculations Impact on energy savings

3.5 Simulations

The CAD-file was imported to the software program Dialux and was used as a foundation for the model. Dialux allowed simple construction for windows, tables, chairs and the luminaries could be imported from manufacturers tab. Reflection factors for ceiling, walls and floor could be adjusted to the correct values for a standard classroom. The constructional dimensions for the room, windows and furniture’s and reflections factors for the surfaces remained constant throughout the simulations for all designs. Only the luminaries were modulated while performing the simulations. The list below summarizes the relevant parameters that’s being evaluated in the simulations, in accordance to table 6 in section 2.11 regarding visual comfort.

• Average illuminance at workplane [E]

• Uniformity at workplane

• Average illuminance at visual task areas [Etask]

• Uniformity at visual task areas [Utask]

• Average illuminance at immediate surrounding areas [Esur r]

IMPROVING VISUAL COMFORT AND ENERGY