After looking at a sample solar file from STRANG in -0:30-0:30 solar and comparing it to a file I've created for the same location (Goteborg Sweden 2015), I've concluded that you are right and that the shift in either the longitude or the time zone should be to the east.
I thought it might be interesting to see how I've confirmed this. In addition to my standard file for Goteborg, I also created two variants where I shifted the longitude east and then west by 7.5 degrees. I then compared the hours of sunrise and sunset by noting the first and last hour with nonzero solar radiation, and plotted those hours where they were different between the STRANG file and my files. For my "standard" file, i.e., with solar as -1:00-0:00, I get the following plot, which indicates there are 133 days where the sunrise hour appears 1 hour earlier and 237 days where the sunset hour appears 1 hour earlier on the STRANG file.
When I shift the building location east by 7.5 degrees, I get the following result. There are much fewer days
where the sunrise and sunset hours are different and the differences are not consistent and most showing sunrise hours earlier and sunset hours later. These differences I attribute to how the solar radiation is modeled, apparently with mine showing more solar at very low sun angles.
When I shift the building location west by 7.5 degrees, I get the following result. Now, almost every day
shows both sunrise and sunset occurring in a later hour.
I originally thought this technique could be used to autodetect the convention being used for solar, but my experience with the Italian IGDG weather files was inconclusive, either because of the lower latitudes, different ways of calculating the sun position ("equation of time"), or physical obstructions if the data were from actual measurements.
Joe
Joe Huang White Box Technologies, Inc. 346 Rheem Blvd., Suite 205A Moraga CA 94556 yjhuang@... http://weather.whiteboxtechnologies.com for simulation-ready weather data (o) (925)388-0265 (c) (510)928-2683 "building energy simulations at your fingertips"
Joe,
That seems to work just as well and is maybe somewhat clearer. Are you sure no other instances than sun position calculation in E+ uses the time zone information?
But you should subtract, not add 0.5 to the time zone. I think it is best shown with the equations used by software like E+.
HourAngle = (15 deg)*(LocalSolarTime - 12 + TimeHourOffset)
LocalSolarTime = LocalTime + (Longitude / (15 deg) ? TimmeZone) + EoT
Longitude and TimeZone are positive in eastward direction. The LocalTime is the time given in the .epw file. HourAngle should be 0 deg at solar noon for data in -0:30-0:30 and -7.5 deg (equals time 11:30) for data in -1:00-0:00 convention.
For data in -0:30-0:30 you need to neutralise that E+ sets TimeHourOffset=-0.5. This is done by adding 7.5 to the Longitude or adding -0.5 to the TimeZone, as can be seen from the equations above.
This half hour discrepancy doesnâ??t make much difference for common simulation purposes. But there are cases when it is important. In hygrothermal simulation solar heat and the hour angle are important parameters. And in some building certifications/regulation system thereâ??s a max limit on allowed solar heat gain through windows, a faulty hour angle calculation would affect this.
Regards,
Lukas