Impact of daylight saving time on lighting energy consumption and on the biological clock for occupants in office buildings

1 Impact of daylight saving time on lighting energy consumption and on the biological clock for occupants in office buildings Author’s name and affiliation: Laura Bellia, Dipartimento di Ingegneria Industriale, Università degli Studi di Napoli Federico II. Ignacio Acosta, Instituto Universitario de Arquitectura y Ciencias de la Construcción, Universidad de Sevilla. Miguel Ángel Campano, Instituto Universitario de Arquitectura y Ciencias de la Construcción, Universidad de Sevilla Francesca Fragliasso, Dipartimento di Ingegneria Industriale, Università degli Studi di Napoli Federico II. Corresponding Author: Laura Bellia, Dipartimento di Ingegneria Industriale, Università degli Studi di Napoli Federico II, Corresponding Author Tel. number: 00393498076838 Email: [email protected] Permanent address: Instituto Universitario de Arquitectura y Ciencias de la Construcción, Universidad de Sevilla, 41012 Seville. Spain. Abstract Nowadays, there is an in-depth debate about the suitability of daylight saving time and its usefulness for energy savings. The shift of one hour during summer time allows a better use of daylighting and a less energy consumption in electric lighting. However, the divergence between solar and local time according to the location, as well as the impact of daylight saving time in the human health, have not been quantified in most of the scenarios, questioning the benefits provided by this time change. The aim of the present study is to determine the impact of daylight saving time on the health of occupants and on the consumption of electric energy for lighting in office buildings. For this purpose, a standard office room is analyzed in eleven representative locations of Europe and for three different time schedules: with the current daylight saving time, with continuous winter time and with continuous summer time. The impact of time schedule on health was quantified by assessing the circadian stimulus (CS) produced by daylight during the first hour of occupancy in the morning, assuming for CS a threshold value equal to 0.3. The influence on energy use was evaluated calculating both Daylight Autonomy and energy consumptions. Despite restricted to a specific case study, results allow to add more elements useful to take a decision about time schedule and confirm that pros and cons in maintaining daylight saving time or abolish it are strongly dependent on the geographical position as well as on the local luminous climate. Keywords: daylighting; daylight saving time; energy savings; circadian stimulus. 2 1. Introduction 1.1. Background During the last century, society has been clearly aware of the need to reduce energy consumption, mainly in buildings, through the proper use of lighting [1], air conditioning [2] and other processes related to human activity. Given this context, the Daylight Saving Time (DST) was proposed by George Hudson in 1895, setting clocks one hour forward compared to the standard time at the beginning of spring and back again in the fall [3]. The aim of this proposal is to promote the synchronization of human activity with daylighting at the peak hours of energy consumption. At present, all member countries of the European Union changes the local time twice a year, although recently the European Parliament has proposed to stop the seasonal clock shift starting in 2021 [4]. Accordingly, each EU member must decide individually about the suitability of applying the DST, whose benefit depends largely on the climate conditions, the variation of the day-night cycle and the deviation of local time (obtained synchronizing the watches of all locations belonging to the same time zone) with respect to solar time (depending on the position of the sun in the sky vault) of a specific location [5]. However, it is worth noting that DST does not only affect energy savings but also has a noticeable influence on the human wellbeing, since the melatonin’s regulation, a hormone secreted by the pineal gland, depends significantly on the light received at the eye’s level [6]. The melatonin suppression is also defined as circadian stimulus (CS) and its regulation synchronizes the circadian rhythm, determining the sleep and alertness patterns, as well as other biological functions [7]. Daylight plays a big role in the circadian rhythms entrainment and the decision to maintain or remove DST or to change the local time could potentially strongly affect circadian human response especially in work environments where activities are carried out starting from obliged times. Both the spectral distribution of the light source and the irradiance received by the observer set the circadian system in a different way they affect the visual one. The suppression of melatonin is more sensitive to visible short-wavelength radiation, with a peak close to 460 nm [8, 9]. Moreover, the visual perception responds in milliseconds to the light received while the circadian response takes several minutes to affect the melatonin suppression. In most cases natural light is an ideal source to promote a suitable circadian entrainment, providing the proper amount, spectral distribution and duration: when available, it should be used. The effects related with the disruption of CS was tested initially in mice and subsequently in humans[10], concluding its relevance for human health. Circadian disruption can also promote depression [11], morbidity[12] and multiple sclerosis [13]. Recently, Rea et al. developed an empirical model for determining the human melatonin suppression, according to the neuroanatomy and neurophysiology of the retina and on results from published psychophysical studies [6]. Despite the fact that the photosensitive retinal ganglion cell (ipRGC) is the main variable in the empirical model, several studies have demonstrated that signals from rods and cones also give information to the biological clock. The method defined by Rea et al. analyzes these 3 multi-channel inputs and quantifies the CS, considering 1 h exposure time with a fixed 2.3 mm diameter pupil. In accordance with the latest research by Figueiro et al. [11] a CS value of 0.3 during the first hours of the day is optimal for the promotion of a good circadian entrainment. There are several metrics in the current context to assess the influence of daylighting and its relationship with electric lighting. Daylight factor (DF) is the most widespread metric for determining the daylight in a space, since it represents the potential of natural light using the worst case scenario, under overcast sky conditions [14]. Many of the research carried out about this topic are based on this metric [15–18]. However, the current tendency follows the application of daylight dynamic metrics, based on climate information, occupancy time and the illuminance requirement[19]. The most common dynamic metric is daylight autonomy (DA), defined by Reinhart et al., which represents the percentage of the occupied time during which an illuminance threshold is reached by daylight alone [20]. Following the trend of the daylight dynamic metrics, the concept of circadian stimulus autonomy (CSA) is proposed, determining the percentage of time throughout the year when a specific threshold of circadian stimulus is met by daylight [21] during the morning. The calculation procedure of CSA is based on DA, since knowing the spectral power distribution (SPD) of the perceived light, the illuminance threshold corresponding to a specific desired CS value can be defined. Usually, CSA is calculated taking into account a CS value of 0.3, as determined before [11], considering the resulting SPD perceived by the eyes of the observer and setting the occupancy hours preferably early in the morning, from 8:00 to 9:00 a.m. Accordingly, this concept can serve to determine the impact of daylighting in the human health. 1.2. Aim and objectives The aim of the presented research is to quantify the impact of DST on the health of occupants in office buildings and its relationship with the energy consumption of the electric luminaires, analyzing the energy saving obtained by the use of the natural light and determining the benefits provided by this source in the biological markers of the human being. Three scenarios are analyzed for eleven locations in Europe (Athens, Berlin, Copenhagen, Lisbon, London, Naples, Paris, Seville, Stockholm, Vienna, Warsaw); the current application of DST, the continuous winter time and the continuous summer time. The locations are selected according to the following different groups:  Southern European cities, with mainly clear sky conditions throughout the year (Athens, Lisbon and Naples);  Central European cities, with intermediate sky conditions (London, Vienna and Warsaw);  Northern European cities, with predominantly overcast skies (Berlin, Copenhagen and Stockholm);  European cities with a high deviation of the local time with respect to solar time (Paris and Seville). 4 The study of the energy savings in electric lighting given by the natural source is determined by means of the analysis of daylight dynamic metrics, previously tested in a real trial under real climate conditions. After the validation of these metrics, the daylight autonomy is used to quantify the percentage of time during which electric lighting is switched off, according to the scenarios described above. Moreover, energy consumption was calculated considering a specific installed power equal to 2.00 W/m2/100 lx. The analysis of biological markers promoted by daylighting are based on the quantification of the melatonin suppression, by means of the CS parameter. The model of circadian light developed by Rea et al. [6] allows to determine the CS value according to an illuminance value and a specific spectral irradiance distribution. The novelty of this study is argued in the analysis of the impact of daylighting on both the energy consumption and on the human health’s response, bringing a new focus about the convenience of the application of DST. 2. Methods The methodology is based on the calculation of a virtual model of a small office with a single window, which is used to assess the impact of three time schedules (Daylight Saving Time—DST—, continuous winter time and continuous summer time) throughout the year both in the electric lighting consumption and in the circadian stimulus of the occupants. In this way, this model is placed in 11 European cities (Athens, Berlin, Copenhagen, Lisbon, London, Naples, Paris, Seville, Stockholm, Vienna, Warsaw) with two orientations (North and South), in order to cover most latitudes, time zones and sky conditions in Europe. Two dynamic metrics are applied to analyze the impact of natural lighting: Daylight Autonomy (DA) allows to evaluate the energy saving, while the Circadian Stimulus Autonomy (CSA) quantifies the circadian entrainment. 2.1. Characteristics of the room model The model under study for dynamic calculations, which corresponds to a small office, is based on a virtual room of 3.0 meters high by 3.0 m wide by 5.0 m deep, with a single long window of 1.2 m high centered on one of its short sides, having a visible transmittance of 0.74. The inner surfaces of the room, considered as Lambertian reflectors, have a reflectance value of 0.86 for the ceiling, 0.72 for walls and 0.43 for the floor, as Figure 1 shows. 5 Figure 1: Calculation model and calculation points location for energy efficiency trials The assessment of dynamic lighting values is performed according to the metric: In the case of DA, horizontal illuminance values were calculated on an array of points on the central axis of the room, located with a spacing of 0.40 m and a height of 0.80 m above the floor, as it can be seen in Figure 1. In this way, it is possible to evaluate the amount of daylight that receives each point of the array of the work plane. In the case of CSA, Figure 2 shows a transverse line of vertical illuminance points 0.50 m from the side wall, also with a spacing of 0.40 m but a height of 1.20 m above the floor and facing the opposite wall, simulating the point of view of a seated occupant. 6 Figure 2: Calculation model and study points location for circadian stimuli trials. To evaluate the CS the spectral reflectance of the inner surfaces of the room is decisive to determine the resulting spectral irradiance at the observer’s eye. In accordance with the calculation model described above, four typical different spectral reflectance distributions are defined for the ceiling, walls, floor and table, as seen in Figure 3. In order to determine the resulting spectral irradiance, it was assumed that for an average position of the observer, the wall surface occupies 40% of the field of view, while the ceiling and the floor fill 10% of the scene. The table and window occupy 20% of the field of view respectively. The average spectral reflectance modifies the received spectrum and therefore the promoted CS. 7 Figure 3: Spectral reflectance values of the inner surfaces of the room model 2.2. Location and orientation of the room In order to determine through dynamic metrics the impact of DST on energy consumption (DA metric) and circadian stimulus (CSA metric), the room under study was located in eleven European cities, as it can be seen in Table 1, with the aim to cover most of the cases of geographical positions and climate conditions. Weather data for all the locations under study were obtained from the EnergyPlus reference [22], which were based on the climate data files from the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) [23] and the Spanish National Institute of Meteorology (AEMET) [24]. Table 1: Geographical parameters and time zone of the cities under study City Code Latitude Longitude Time zone Centre of the time zone Distance from the center of the time zone Percenta g e of hours with Direct sun Clear skies Overcast skies Athens ATH 37° 58' 23° 42' E GMT +2 30° 0’ -6° 17' 80% 26% 39% Berlin BER 52° 31' 13° 24' E GMT +1 15° 0’ -2° 35' 58% 8% 66% Copenhagen COP 55° 40' 12° 33' E GMT +1 15° 0’ -3° 26' 62% 7% 64% Lisbon LIS 38° 43' -9° 7' O GMT +0 0° 0’ -9° 7' 95% 22% 15% London LON 51º 30’ -0º 7’ O GMT +0 0° 0’ -0º 7’ 72% 25% 36% Naples NAP 40°51' 14°16' E GMT +1 15° 0’ -1° 43' 77% 12% 48% Paris PAR 48° 51' 2° 20' E GMT +1 15° 0’ -13° 39' 62% 7% 66% Seville SEV 37° 22' -5° 58' O GMT +1 15° 0’ -20° 1' 81% 31% 32% Stockholm STO 59° 19' 18° 3' E GMT +1 15° 0’ +3° 3' 61% 8% 61% Vienna VIE 48° 12' 16° 22' E GMT +1 15° 0’ +1° 22' 61% 11% 63% Warsaw WAR 52° 13' 21° 0' E GMT +1 15° 0’ +6° 0' 61% 6% 65% Figure 4 shows the day-night cycle in each of the cities under study according to the climate data base of EnergyPlus, both on the day of the year with the highest number of hours of daylight (summer solstice) and on the day with fewer hours of daylight 8 (winter solstice), as well as the equinox sunrise and sunset. Vertical lines in Fig.4 represent when the solar noon occurs in each location. As it can be seen, latitude of the site greatly affects the number of daylight hours and the occurrence of the hours of sunshine throughout the whole year, whereas longitude and specifically the distance from the considered location to the center of its corresponding time zone moves significantly daylight hours throughout the day. As a consequence, daylight hours in winter are in most cases distributed almost symmetrically with respect to the noon, with the exception of locations far from the center of the time zone as Seville, Warsaw, Paris and Lisbon. This is particularly noticeable for the cases of Paris and Seville. During summer, thanks to the delay produced by the DST, the asymmetry is corrected, as in Warsaw, or is amplified, as in Seville or Paris. By comparing Seville and Paris, it can be observed that sunset occurs nearly at the same time, whereas in Seville, at a lower latitude, sunrise occurs later than in Paris. Figure 4: Day-night cycle with DST of the cities under study in summer and winter solstices For each of the locations, north and south orientations were used for calculating both the DA and CSA promoted by lighting, given that both orientations can show the best and worst case scenarios for the use of natural light, when east and west orientations usually provide intermediate scenarios [19]. For the CSA assessment, the daylight SPD is necessary: for this reason, a typical average sky SPD was built, based on the climatic characteristics of each location and on their latitude. Figure 5 shows the average SPD obtained for the sky of three types of location regarding to the latitude (southern Europe, middle Europe and northern Europe), according to statistical data on the percentage of overcast skies [22], as well as the spectral 9 distribution of the 6500 K sky light (CIE D65) presented as average SPD for northern Europe. As it can be seen in Figure 5, the SPD in northern locations has a distribution close to that of a typical overcast sky (CIE Standard Illuminant D50 distribution), while in middle locations the curve of the SPD is similar to the CIE D65 distribution. Finally, the locations in the southern Europe, mainly with clear skies, show a SPD slightly higher in the range of the short wavelength than the CIE D65 distribution. Figure 5: Daylight average spectral power distribution of the locations under study, according to statistical data of Energy Plus and percentage of overcast skies throughout the year. 2.3. Scenarios under consideration The performance of the DST according to DA and CSA metrics was evaluated through its comparison with other two annual schedule scenarios, in which there were no spring or autumn time changes. These scenarios are:  Scenario 1 (defined from now on “winter time”) for which the standard time is applied for the entire year, without considering the 1-hour shift during spring and summer typical of DST;  Scenario 2 (defined from now on “summer time”) for which the 1-hour shift that in DST is applied exclusively in spring and summer is applied for the entire year. In this way, this set of scenarios allows the evaluation of the impact of the time change according to its closest annual schedules —winter time and summer time throughout the year—. 2.4. Parameters of the modelling tool The lighting dynamic calculations of this research were performed using DaySim 3.1 simulating tool, which is based on the Radiance engine of the Lawrence Berkeley National Laboratory, including the definition of the current metrics regarding the actual sky definitions [25]. This tool was validated by previous studies [26–30] through comparison with the CIE test cases [31]. 16 Figure 9: Section views of the calculation models in southern cities, according to DST, winter time and summer time, showing DA for 500 lx in the horizontal plane, the equivalent energy consumption and CSA for 200 lx in the vertical plane. 17 Figure 10: Section views of the calculation models in central cities, according to DST, winter time and summer time, showing DA for 500 lx in the horizontal plane, the equivalent energy consumption and CSA for 200 lx in the vertical plane. 18 Figure 11: Section views of the calculation models in northern cities, according to DST, winter time and summer time, showing DA for 500 lx in the horizontal plane, the equivalent energy consumption and CSA for 200 lx in the vertical plane. 19 Figure 2: Section views of the calculation models in cities with a noticeable divergence between local and solar time, according to DST, winter time and summer time, showing DA for 500 lx in the horizontal plane, the equivalent energy consumption and CSA for 200 lx in the vertical plane Moreover in Table 3 for each location and for both orientations the DA and CSA values obtained by applying the DST are reported for three points: the point closest to the window (0.50 m from the façade), the point equally distant from the window and the rear wall (2.50 m from the window) and the point closest to the rear wall (4.50 m from the window).With respect to these values, considered as base case and reported in the first row for each city, the percentage differences obtained by considering the winter time throughout the entire year and the summer time in the entire year were reported in the second and third row. 20 Table 3: DA and CSA for the studied locations in accordance with the distance from the façade and the application of DST, showing the relative difference for the scenarios of winter time and solar time DA (500 lx) CSA (200 lx) North South North South 0.50 m 2.50 m 4.50 m 0.50 m 2.50 m 4.50 m 0.50 m 2.50 m 4.50 m 0.50 m 2.50 m 4.50 m Athens DST 80.41 71.58 59.48 80.67 73.56 66.02 100.00 97.41 90.13 88.98 80.59 82.69 Winter Time -5.5% -5.2% -4.7% -5.5% -5.2% -4.3% 0.0% 1.0% 2.7% 3.8% 7.8% 6.8% Summer Time 1.0% 0.3% 0.0% 1.1% 0.4% 0.2% -22.4% -28.2% -29.0% -30.0% -29.2% -29.7% Berlin DST 72.93 52.80 37.77 73.22 56.83 45.79 88.28 73.24 64.50 63.60 54.47 56.26 Winter Time -5.7% -5.9% -5.5% -5.7% -5.8% -5.1% 0.6% 4.6% 7.8% 8.5% 14.9% 13.3% Summer Time 0.7% 0.0% 0.0% 0.7% 0.1% 0.0% -21.1% -15.3% -15.7% -16.6% -15.7% -15.8% Copenhagen DST 72.49 52.39 37.54 73.13 59.04 49.07 80.68 70.75 61.17 60.44 53.77 55.36 Winter Time -5.7% -6.2% -5.9% -5.7% -5.7% -5.2% 1.5% 5.4% 10.9% 12.5% 16.0% 15.2% Summer Time 0.6% 0.3% 0.0% 0.6% 0.3% 0.2% -17.7% -13.6% -9.8% -11.0% -9.5% -9.4% Lisbon DST 78.07 68.77 56.32 78.54 70.78 60.89 99.55 98.21 92.91 94.22 89.88 90.90 Winter Time -5.7% -5.6% -5.5% -5.6% -5.4% -5.1% 0.1% 0.3% 1.6% 2.0% 3.7% 3.1% Summer Time 1.5% 0.7% 0.4% 1.6% 0.8% 0.6% -22.6% -24.6% -24.6% -23.4% -25.3% -25.0% London DST 80.15 66.06 48.53 80.37 69.44 58.22 98.47 91.28 79.47 80.49 70.79 72.96 Winter Time -5.3% -4.9% -4.6% -5.4% -5.1% -4.1% 0.6% 2.2% 4.9% 4.8% 9.2% 8.1% Summer Time 0.8% 0.2% 0.0% 0.9% 0.5% 0.2% -27.4% -27.3% -23.2% -24.2% -24.0% -23.8% Naples DST 77.41 67.81 55.21 77.50 70.19 62.48 100.00 98.88 94.13 93.23 85.82 87.36 Winter Time -5.8% -5.9% -5.5% -5.8% -6.0% -5.5% 0.0% 0.2% 0.9% 1.2% 3.3% 2.6% Summer Time 1.6% 0.4% 0.1% 1.7% 0.7% 0.3% -13.5% -22.5% -26.4% -26.5% -26.3% -26.5% Paris DST 77.54 58.68 42.73 77.87 61.99 50.24 73.53 62.13 51.66 49.14 39.24 41.70 Winter Time -4.8% -4.1% -3.0% -4.8% -3.9% -2.1% 6.0% 11.8% 20.6% 24.3% 41.7% 35.6% Summer Time 0.1% 0.0% 0.0% 0.1% 0.0% 0.0% -20.6% -16.3% -12.5% -12.0% -9.4% -9.8% Seville DST 84.32 73.97 61.20 84.73 75.82 67.67 69.25 56.51 41.79 36.56 23.98 27.08 Winter Time -3.4% -2.0% -1.0% -3.4% -1.4% -0.1% 15.8% 29.4% 57.0% 75.2% 137.7% 117.5% Summer Time 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% -28.8% -25.3% -21.6% -22.4% -20.1% -20.4% Stockholm DST 67.46 47.72 33.28 68.53 54.60 45.37 80.27 70.91 62.87 60.44 53.77 55.36 Winter Time -6.2% -6.8% -7.0% -6.2% -6.4% -6.2% 1.2% 4.0% 6.1% 12.5% 16.0% 15.2% Summer Time 0.7% 0.3% 0.0% 0.7% 0.4% 0.2% -15.4% -12.2% -9.4% -11.0% -9.5% -9.4% Vienna DST 73.39 56.90 41.70 73.66 60.02 49.87 98.37 87.20 75.67 74.55 66.09 68.26 Winter Time -5.9% -6.1% -5.9% -5.9% -6.0% -5.6% 0.1% 1.9% 4.8% 5.1% 9.2% 7.9% Summer Time 1.1% 0.4% 0.0% 1.1% 0.6% 0.2% -22.0% -21.6% -18.1% -18.3% -18.0% -18.2% Warsaw DST 70.25 52.11 37.83 70.55 55.88 46.17 82.09 73.63 70.27 74.71 66.35 68.26 Winter Time -6.2% -6.7% -7.1% -6.3% -6.6% -6.5% 4.3% 4.2% 4.5% 5.0% 9.0% 7.7% Summer Time 1.3% 0.3% 0.0% 1.4% 0.6% 0.3% -17.0% -15.7% -15.1% -16.8% -16.8% -15.9% 4. Analysis of results 4.1 Daylight Autonomy and energy consumptions Considering daylight availability, despite the aim of this research is to assess the differences in terms of potential energy saving and circadian stimulus in a typical room office, by adopting different times (DST, Winter time and Summer time) in several European locations, it’s worth considering also the differences among the chosen sites. These are due not only to the differences of latitudes, but also to the specific climatic characteristics. For all the diagrams in the Figures from 9 to 12 it can obviously be inferred that DA values increase on decreasing the distance from the window. Moreover, in most cases, higher values are associated to lower latitudes. For example, considering the DST scheduling at a distance of 0.5 m from the window DA values are generally around 80% for southern cities (Figure 9), around 70-75% for the central ones, with London reaching about 80% 21 for each orientation, (Figure 10) and around 70% for northern ones (Figure 11). As it can be expected, with the same conditions and for each location, results obtained for the south orientation are always greater than those calculated for the north one. This is particularly evident in the points in proximity to the rear wall, whereas for the points closer to the window DA values, for each case, show neglectable differences on changing the window orientation. The differences among the sites with very different latitudes are greater for the points closer to the rear wall, whereas they are slightly attenuated near the window. Stockholm is characterized by the lowest DA values, but for the south exposure the results are very close to the Warsaw’s ones. The highest DA values are observed in Sevilla, and this can be explained not only by the low latitude (close to the Athens and Lisbon ones) and predominantly clear skies, but also by the fact that Seville lies at a distance of about 20° West from the center of its time zone. As it can be inferred in Table 3 in every case, comparing DST schedule with winter time it can be noticed that DA is always reduced, being the lowest variation in Seville, with a maximum reduction of around 3.4% for both orientations in the points closest to the window. The highest reductions are observed in Stockholm and Warsaw, with values ranging from 6,19% (Stockholm, South orientation, the point closest to the window) to 7.14% (Warsaw, North orientation and the point closest to the rear wall). For Berlin, Copenhagen Lisbon, Naples and Vienna the reductions are always comprised between around 5.0% and 6.0%, whereas for Athens and London the variations range from around 4.0% to nearly 6.0%. In any case, percentage differences seem not to be linked neither to the specific window orientation nor to the distance of the calculation point from the window. The percentage differences obtained by comparing DA values in the base case (DST), with the daylight saving time applied for the entire year (“summer time”) are negligible for most cases. The highest percentage difference is +1.72% in Naples for the point closest to the window and most of the differences are lower than 1.0%. So, it can be inferred that, as far as DA is concerned, the most favorable conditions are DST and “summer time”. As reported in the graphs of Figures 9-12, considering the correspondence between DA and energy consumptions (the yearly percentage of hours during which the lighting system is turned on is the complementary to 100% of DA), it can be concluded that the winter time application for all the year determines an increase of energy consumptions, whereas the summer time a reduction. Specifically, the increase observed for winter time is common to all cities, particularly significant for Stockholm and Warsaw and it is on the average around 5%. On the other hand, the increase due to summer time application is negligible for all the cities (around 1%). 4.2 Circadian stimulus As it can be inferred by observing Figures from 9 to 12, differently from the DA results, evaluated for the whole period of occupancy, CSA values are affected by significant variation when maintaining the same time profile, but on changing the window exposure. Indeed, for the north exposure and for all the locations, CSA values are always decreasing on increasing the distance 22 from the window, being the specific values and slopes depending on the location. Conversely, for the south orientation CSA decreases on increasing the distance from the window, attaining a minimum value at 3.30 m and then increasing again with smaller increments up to the rear wall. Moreover, in all cases, south orientation values are always lower than the north orientation ones, and the differences are smaller in the points closer to the rear wall, in some cases attaining their maximum value in the center of the room. These differences at the center of the room are more or less significant depending on the cities. Among southern ones, in Lisbon they are quite small (7 to 8% in the central part of the room depending on the time profile) whereas they are more consistent in Athens (ranging from 11% to 17%) and Naples (ranging from 10% to 14%). As for the central cities the lowest differences are observed for Warsaw (ranging from 5% to 7%), whereas they reach significant values for London and Vienna being around 13-20%. As for the northern cities, the differences are around 15-18% in Berlin and around 11 to 17% in Copenhagen and Stockholm. Finally, in Paris and Seville differences are really noticeable for DST (22% in the French city and 33% in the Spanish one), they reduce in summer time to 16% and 23% respectively, being equal to 4% and 6% in winter time. The differences due to the orientation are also dependent on the time profile, being generally greater for DST and summer time. Considering the differences among the calculation points due to the distance from the window, in the southern cities it can be observed that when the room is north-oriented the point closest to the window is always characterized by 100% CSA both with DST and winter time, whereas in the rear part of the room, values equal to at least 90% are achieved. Conversely, when the window is south-oriented, in the point next to the window CSAs range from 90% to 96% and decrease to 83% to 94% in the farthest point depending on the city and the time profile. For both orientations the summer-time turns out to be the most disadvantaged one ranging the CSAs from 58% to 86% depending on the orientation and the city. In central cities, considering the north orientations, the differences due to the distance from the window are more significant for DST and winter time and for London and Vienna (around 15-25% depending on the city and on the scheduling), whereas they are lower for summer time (never higher than 15%) and for Warsaw (among 7 and 10% depending on the scheduling). As for the south orientation CS curves assume a concave trend, being the CS values really similar at the extreme parts of the room. For northern cities the most disadvantaged profile is always the summer-time with values ranging from 50% (in the rear part of the room) to 68% (in the frontal one), whereas DST and winter time show higher values especially in the case of north orientation with CSAs equal to around 80% (even 90% in Berlin case) next to the window and ranging from 61% to 73% next to the rear wall. In general, the lowest CS values are related to Paris (ranging from 38% to 49%) and Seville (from 23% to 37%). This can be attributable to the fact that both of them, irrespectively of the different latitude, are located at a distance greater than 15° from the center of their time-zone towards west, so sunrise occurs later than other locations at the same latitude. The most disadvantaged is Seville, because during summer the sunrise occurs later than in Paris, given the lower latitude. Moreover, during summer, the first hours of the day are characterized by a solar azimuth greater than 90°, consequently the northern part of the 23 sky is brighter than the southern one: this explains also the lower values for the south exposure even for the other sites and also the fact that the differences due to orientation are greater for DST and summer time: indeed for both of them, during summer, the sunrise is anticipated with respect to the solar (winter) time. As it can be pointed out by observing the results in Table 3, CSA values for winter time are always slightly greater than those calculated with DST (maximum values registered for the point close to the rear wall in Copenhagen and equal to about 11% with north orientation and to 15% with the south one), except for Paris and Seville where increments are comprised between around 6% and 21% (Paris, north exposure), 24% and 42% (Paris south exposure), 16% and 57% (Seville north exposure) and 75% and 138% (Seville south exposure). The worst condition is obtained by applying the summer time, with decrements in CSA mostly greater than 10% and specifically exceeding 20% in the cases of Athens, London, Seville and Lisbon. 5. Conclusions The goal of this paper is to investigate the consequences connected to the recent proposal of the European Parliament to stop the application of the Daylight Saving Time (DST), i.e. the proposal to avoid the seasonal clock shift allowing to gain a supplementary daylight hour during summer evenings. The benefits connected to the application of DST depend on the climate conditions, the variation of the day-night cycle and the deviation of local time (obtained synchronizing the watches of all locations belonging to the same time zone) with respect to solar time (depending on the position of the sun in the sky vault) of a specific location. Originally, the application of DST was proposed for energy saving purposes: shifting the social life according to the natural seasonal daylight rhythm determines a better exploitation of the natural source and a reduction of the operative hours of lighting systems. However, studies about non-visual effects of light have largely demonstrated that human biological clock synchronization is strictly dependent on the very daylight natural rhythm. As a consequence, DST application has an influence on the regulation of people circadian rhythms. Based on these premises, the paper aimed at evaluating the impact of DST application on both energy saving and regulation of the circadian rhythms by calculating two metrics Daylight Autonomy (DA) and Circadian Stimulus Autonomy (CSA) in eleven European cities, characterized by different weather conditions (Athens, Berlin, Copenhagen, Lisbon, London, Naples, Paris, Seville, Stockholm, Vienna, Warsaw). In order to do that, three different time schedules were used: DST, winter time for the entire year and summer time for the entire year. For each time schedule, dynamic daylight simulations were performed referring to a simple office, alternatively facing south and north orientation. As far as DA is concerned, the worst condition in the office is obtained by applying the winter time throughout the year. However, the decreasing related to the DST currently adopted, is not relevant (with most of percentage reductions lower than 6%). Small 24 differences are observed among analyzed cities. Considering the central point of the room, the decrement of DA is around 5.26.0% for southern cities, 4.9-6.7% for central ones, 5.7-6.8% for northern ones and 1.4-4.1% for Paris and Seville. On the other hand, applying the 1-hour shift for all the year (summer time), does not affect the DA values for the considered office in all the selected cities: considering the central point of the room, the DA increases are indeed always lower than 1%. Consequently, as far as only the electric consumption due to lighting is considered, the best solution is to adopt the DST as it is currently used in Europe. It must be noted that this result is limited only to the office use and does not considers other energy uses. As regards implications in terms of circadian rhythms regulation, the application of the winter time guarantees an increase of the CSA values. In this case the differences are significant among cities, indeed the increments of CS values range from 0.2 to 7.8% for southern cities, from 1.9 to 9.2% for central cities, from 4.0 to 16.0% for northern ones and from 11.8 to 137.7% for Paris and Seville. So, the CS increment grows on latitude increasing and assumes the highest values for the cities characterized by a noticeable divergence between local and solar time. On the contrary, the use of summer time, determines decrements in CSA mostly greater than 10% and specifically exceeding 20% in the cases of Athens, London, Seville and Lisbon. Also, in this case, the decrements depend on the latitude, being higher for southern cities and the maximum values are registered for Paris and Seville. In summary, regarding the selected case study the elimination of DST, by applying the winter time all the year, would determine a slight increment of energy consumptions for all the cities corresponding to an increment of CSA values more or less significant depending on the considered cities and particularly favorable for northern ones and for cities characterized by a noticeable divergence between local and solar time. On the contrary, the use of summer time would determine decrements of CSA values for all the cities and particularly disadvantageous for the southern ones but, at the same time this decrement would be not balanced by a convenient increment of energy saving. As it was continuously stressed in the paper, the most significant effects of DST application in terms of circadian rhythm regulation were observed for the cities characterized by a noticeable divergence between local and solar time, demonstrating that the more the social life rhythm is adapted to the daylight one, the more benefits are achieved considering wellness implications. Acknowledgements The collaboration between the authors arises by means of research stays funded by the government of Spain through the research project: Efficient design for biodynamic lighting to promote the circadian rhythm in shift work centers (Ref BIA2017-86997-R). The authors wish to express their thanks for all the technical and financial support provided. The authors also wish to thank the Junta de Andalucía and the University of Seville for providing the test cell facilities which made it possible to validate the metrics used in this research. 25 References 1. Acosta, I.; Campano, M. A.; Molina, J. F. Window Design in Architecture: Analysis of Energy Savings for Lighting and Visual Comfort in Residential Spaces. Appl. Energy, 2016, 168, 493–506. https://doi.org/10.1016/j.apenergy.2016.02.005. 2. Domínguez-Amarillo, S.; Fernández-Agüera, J.; Campano, M. A.; Acosta, I. Effect of Airtightness on Thermal Loads in Legacy Low-Income Housing. Energies, 2019, 12 (9), 1–14. https://doi.org/10.3390/en12091677. 3. Kudela, P.; Havranek, T.; Herman, D.; Irsova, Z. Does Daylight Saving Time Save Electricity? Evidence from Slovakia. Energy Policy, 2020, 137 (April 2019), 111146. https://doi.org/10.1016/j.enpol.2019.111146. 4. European Commission. European Commission Welcomes the Parliament’s Endorsement to Put an End to Seasonal Clock Changes. Technical Report, European Commission. Daily News. Brussels 2019. 5. Havranek, T.; Herman, D.; Irsova, Z. Does Daylight Saving Save Electricity? A Meta-Analysis. Energy J., 2018, 39 (2), 35–61. https://doi.org/https://doi.org/10.5547/01956574.39.2.thav. 6. Rea, M. S.; Figueiro, M. G.; Bullough, J. D.; Bierman, A. A Model of Phototransduction by the Human Circadian System. Brain Research Reviews. December 2005, pp 213–228. https://doi.org/10.1016/j.brainresrev.2005.07.002. 7. Leslie, R. P.; Radetsky, L. C.; Smith, A. M. Conceptual Design Metrics for Daylighting. Light. Res. Technol., 2012, 44 (3), 277–290. https://doi.org/10.1177/1477153511423076. 8. Brainard, G. C.; Hanifin, J. P.; Greeson, J. M.; Byrne, B.; Glickman, G.; Gerner, E.; Rollag, M. D. Action Spectrum for Melatonin Regulation in Humans: Evidence for a Novel Circadian Photoreceptor. J. Neurosci., 2001, 21 (16), 6405–6412. https://doi.org/21/16/6405 [pii]. 9. Thapan, K.; Arendt, J.; Skene, D. J. An Action Spectrum for Melatonin Suppression: Evidence for a Novel Non-Rod, Non-Cone Photoreceptor System in Humans. J. Physiol., 2001, 535 (1), 261–267. https://doi.org/10.1111/j.14697793.2001.t01-1-00261.x. 10. Bullough, J. D.; Rea, M. S.; Figueiro, M. G. Of Mice and Women: Light as a Circadian Stimulus in Breast Cancer Research. Cancer Causes and Control. May 2006, pp 375–383. https://doi.org/10.1007/s10552-005-0574-1. 11. Figueiro, M. G.; Steverson, B.; Heerwagen, J.; Kampschroer, K.; Hunter, C. M.; Gonzales, K.; Plitnick, B.; Rea, M. S. The Impact of Daytime Light Exposures on Sleep and Mood in Office Workers. Sleep Heal., 2017, 3 (3), 204–215. https://doi.org/10.1016/j.sleh.2017.03.005. 12. Radetsky, L. C.; Rea, M. S.; Bierman, A.; Figueiro, M. G. Circadian Disruption: Comparing Humans with Mice.