All scenarios used 7:00 o’clock as waking time to determine daylight use. For instance, in the status quo, the Sun rises in Istanbul at 06:09 on average in March and sets at 18:15, which makes 12 hours and 6 minutes of existing daylight (te = 12 hours and 6 minutes). However, if waking time is taken as 7:00, daylight use decreases to td = 11 hours 15 minutes. The remaining 51 minutes is spent in sleep. Relative error is calculated as 7.02% using these values in Equation 1, for March. In other words, 7.02% of daylight was wasted in March. For the month of January in Istanbul, the Sun rises at 7:21 on average and sets at 17:05. Existing daylight is te = 9 hours and 44 minutes. On the premise that waking time is 7:00, electricity will be consumed for 21 extra minutes of lighting. Even though all existing daylight was used in January, the extra 21 minutes of morning electricity use was reflected in the calculation of relative error, and therefore daylight use was accepted to be td = 9 hours and 23 minutes. These values were used in Equation 1 and relative error for January was calculated as 3.94%. The same approach was used for all months.
For the month of March in Istanbul, in the scenario 3 where the clocks move 30 minutes forward with single DST, the Sun rises at 6:39 and sets at 18:45. Although the existing daylight is still the same with the status quo, which is te = 12 hours and 6 minutes, daylight use rises from td = 11 hours and 15 minutes to td = 11 hours and 45 minutes, if 7:00 is considered as waking time. If we compare this scenario to the status quo, the sun will have risen not 51 but 21 minutes before waking time. Thus, daylight would be used for 30 minutes longer in March, it would be 30 minutes closer to daylight point. Using these values in Equation 1, the relative error is calculated as %2.89. For the month of January in Istanbul, in scenario 3, the Sun rises at 7:51 on average and sets at 17:35. Naturally existing daylight is still te = 9 hours and 44 minutes. As the Sun will have risen 51 minutes after waking time, an extra 51 minutes of electricity will have been used and td = 8 hours and 53 minutes. Using these values, Equation 1 calculates relative error for January as 8.73%. Using this approach for all months, relative error was calculated for Scenario 3. While calculating relative error, it was assumed that people would not change their habits in all scenarious.
Fig. 5 shows existing daylight, status quo, and various scenarios for all months for Istanbul, the representative of western cities. Fig. 6 shows the same information for Erzurum, the representative of eastern cities.