All,
We were looking at weather data files for Sydney Australia,
(Sydney.Airport.947670), and we found that there is a of about an hour for
some of the data in AUS_NSW.Mascot-Sydney.Airport.947670_RMY versus
AUS_NSW.Sydney.947670_IWEC.
Please see the attached jpg. Looking at the direct normal radiation and
radiation-rose you can see that sun-rise and sun-set in RMY file happening
earlier. When combined with solar position calculations based location and
time this makes the sky to appear to be skewed, with more radiation from
the east. We found a similar pattern in AUS_NSW.Sydney.947680_RMY weather
file.
Which weather file is the one that you suggest us to use for the analysis?
Is there a known difference between the RMY and IEWC standard that would
account for this?
Is it because Sydney is near the eastern edge of its time zone, and the RMY
standard is defined in terms of actual local time rather than ?solar time?
and the IEWC standard were defined in terms of solar time rather than local
time?
Regards,
Mostapha
PS: I uploaded the jpeg file here because the list does not allow sending
me the email with the attachments (>200k): (
https://dl.dropbox.com/u/16228160/Sydney_Weather_Data_Comparision.jpg)
I took a look at an old atlas and Sydney's near the middle of its time zone. Please keep
in mind that the solar data in the vast majority of weather files is modeled, i.e.,
calculated from other parameters or satellite imagery, so they're quite dependent on the
solar angle calculations. I know that's the case with the IWEC, not positively sure about
the RMYs, though.
I'm attaching a scratch version of the weather file for Sydney I created for the ASHRAE
IWEC2 set (file sent in separate e-mail to just Mostapha, I think I'm allowed to do
that). Please look at that for a comparison, and let me (and others) know what you find.
The file is in text, and the columns are clearly identified.
Joe Huang
Dear All,
I added the weather data that I received from Weather Analytics in the 3rd
row, and the data from White Box Technologies (IWEC2) in the 4th.
Here is the graphs:
https://dl.dropbox.com/u/16228160/Sydney_Weather_Data_Comparision_II.jpg
And here is the graphs with normalized scales for row 3 and row 4:
https://dl.dropbox.com/u/16228160/Sydney_Weather_Data_Comparision_II_Normalized.jpg
The pattern of the radiation rose for both graphs are mostly similar to
IWEC graph, however the wind-rose for Weather Analytics file is slightly
skewed toward north-west. As a conclusion there should be an issue with the
RMY file.
I also did a comparison for average global and direct normal radiation
between the three weather files (IWEC, WeatherAnalytics, IWEC2). Average
global radiation in WeatherAnalytics data is 9.1% more then IWEC2. This
number is 20.7% more for direct normal radiation.
Here are the graphs for monthly and annual comparison:
https://dl.dropbox.com/u/16228160/radiationComparison_Charts.jpg
Unfortunately I don't have the measured data for comparison.
Excel spreadsheet is attached to this email.
Regards,
Mostapha
Mostapha,
So what was your conclusion about the hourly profile in the solar radiation, which was
your original question?
I contend that since Sydney is close to the standard meridian for its time zone (150
degrees east), the solar profile should be symmetrical around noon.
On the total amounts of solar radiation, I have no basis to judge between the three. Did
you leave off the RMY for some reason? Since these are all "typical year" weather files,
they are likely to be different months from different years. Overall, it does seem that
the WX is showing more solar than the IWEC or IWEC2, which are quite close. The larger
variation in direct normal is understandable, since all the models (as well as
measurements) show that when total goes down, the fraction direct goes down even more.
The only way to evaluate these weather files is to find some actual measurements, even if
it's just of monthly or year totals. Otherwise, we can only say that there are differences.
Joe Huang
Hi Joe,
For or purpose of designing the mass of the building at the early
conceptual stage of the design, we are more concerned with the relative
magnitudes of solar radiation. This is not to say that we are entirely not
concerned with the solar radiation values, but that is less of an
importance at this stage.
Since the RMY file shows a skewed radiation-rose different from all the
other files we decided not to use that file for the studies. I also agree
with your point about the location of Sydney. I double checked and it is
near the center of UTC+10: (http://www.timeanddate.com/time/map/)
Regards,
Mostapha
A quick and, but not definitive response. There are known problems with the RMY files for Australian which I believe are left unresolved at this time as the government ran out of funding to correct the problems. And yet we have them ratcheting up the energy provisions within our construction code and expecting sensible if not deliverable results from JV3 simulations (the equivalent of a 90.1 energy cost budget simulation)!
So today we still use Test Reference Year (TRY) Files that can be bought from ACADS-BSG in raw data form. The BIN files on DOE-2.COM are built off of these TRYs.
Aside from that the radiation data itself is unsymmetric in Sydney. I have not done a definitive analysis but we tend to see clearer mornings in the weather file leading to higher DNI and higher overall global horizontal in the morning half of the hemisphere than the afternoon half. Over the course of the day our intense sun (10% more radiation or so in a southern hemisphere summer due to the elliptical orbit we have around the sun) leads to evaporation and cloud building up that lowers both DNI and global horizontal with all else equal. I thought it was a weather file problem years ago but satisfied myself it wasn't in the end. This wouldn't explain observing the sun rising earlier or later relative to the sunset if you are otherwise in the middle of the time zone...
Sorry I can't provide more background on the Aussie RMY's but I need to get off to work.
G
Graham,
Thanks for the information. I downloaded the ACADS-BSG weather file and also ran some
statistics on it.
I also went on the web and within a few minutes found several sources for solar radiation
in Sydney, which seems
to average between 16-17 megajoules /m2 or 4444-4722 Watts/m2 per day. If anyone wants
to devote a few
hours to this, I'm sure one can get quite a good sense of the measured solar radiation in
Sydney.
I also recalculated the totals from ACADS, IWEC, and IWEC2, and think that Mostapha made a
mistake on the
IWEC2 total horizontal, because what I got was 4519 rather than the 4180 reported by
Mostapha :
ACADS
IWEC IWEC2 WX (as calculated by Mostapha)
(all in W/m2 per day)
Total Horizontal 4386 4526
4519 4940
DIrect Normal 4442 3979
3628 4802
Based on what I've seen so far, I would say that the IWEC and IWEC2 solar are in the
middle of the measured data, the ACADS is somewhat low, and the WX seems somewhat high.
Joe Huang
My fault. When I recalculated the IWEC2 solar using awk, I got the same numbers as
Mostapha, rather than the higher number I got using Excel. So now, it would seem that
IWEC is right in the middle, ACADS and IWEC2 are low, and WX is high.
Joe Huang
Thank you Joe and Graham for the information.
I'm curious to know where I made the mistake though. Isn't the number equal
to the average of the column L (total diffuse radiation) * 24 hours = 174.15
* 24 hours = 4179.6? Probably I'm missing something here.
Thanks,
Mostapha
Sorry Joe! I didn't see your last email... Ignore the previous email.