G'day everyone!
Very often I receive results from my solar exposure analysis that I believe cant't be realistic. So I set up a very simple model. 1m² facing east, 1m² facing west, 1m² horizontal.
With that model and the brightest day in Sydney I experimented a bit what is influencing the result and what is not. I chose the brightest day of the year since the irradiation on the eastern and western surface should be very similar with the given weatherdata.
The results of the first model - Ost-West.eco - seem very strange to me:
east: 1557 Wh/m²
west: 3596 Wh/m²
horizontal: 8269 Wh/m²
after playing around a bit I got better results (see 080702Ost-West.eco):
east: 3490 Wh/m²
west: 3692 Wh/m²
horizontal: 9606 Wh/m²
All results are from the solar exposure analysis but with solar analysis I had similar problems in the beginning.
Now, in the beginning I thought that after doing an inter-zonal adjacencies calculation the results became realistic but when I tried to do it again in "Ost-West.eco" I failed. Is there a difference between the shading mask calculated during an inter-zonal adjacencies calculation and the one calculated during a detailed shading calculation? I didn't think so.
Can somebody tell me what I did right by chance?
What steps do I need to take in "Ost-West.eco" to get the results from the other model? This is an easy model but I couldn't always check on correctness with my other models.
Thanks for help,
Thomas
| Attachment | Size |
|---|---|
| Ost-West.eco | 114.8 KB |
| 080702Ost-West.eco | 114.8 KB |
Hi Thomas,
Hi Thomas,
Looking at your model and doing the calculations, the reason you get such a differential between the two orientations is entirely down to the weather data you are using.
Because of cloudiness, there is simply more cumulative solar radiation available in the afternoon than there is in the morning. You can see this by the area of blue compared to the area of red in the above image. If you try a few different days, you get different results.
The key here is to test it over a longer period. However, even here, testing from 8am to 7pm means that the daily exposure is not symmetrical about solar noon, so one side is tested for longer than the other side. Next there is the issue of actual longitude versus reference longitude where the time zone is taken. This will mean that solar noon does not occur at exactly local noon. A more symmetrical period would be 7am to 5pm, which over the whole year gives 600500Wh/m2 against 728271Wh/m2.
This brings us to the characteristic of this particular climate. Being on the east coast, cloud conditions and atmospheric turbidity in the morning will always be very different to the afternoon / evening, resulting again in significant differences in solar availability.
Unfortunately, real weather data gives 'imperfect' results.
Hope that helps...
Kind regards,
Andrew
Dr. Andrew Marsh
Square One research
http://squ1.com
Hey Andrew, I'm aware of
Hey Andrew,
I'm aware of the effect that clouds and the rest of the weather data in general have on the results. that is why I chose december 17th, brightest day of the year, for my calculations.
If you check the weather during that day you'll see that the distribution of the insolation is about symmetrical over day.
Question about the location: I know that the sun doesn't reach its highest altitude at noon if the location is not on the standard meridian of the applicable time zone. But, when creating wea files from meteonorm or taking one that's offered for download is it right that I always use the actual location - first when creating the meteonorm file, then again in the weather tool when transforming the weather data and again in my simulations in ecotect? Or is it possible to over/ underestimate the shift in time by doing so?
Either way, there is always one hour of the day during which the sun passes south. During that hour the sun sends some radiation onto the eastern and some onto the western surface. For calculation purposes, is the position of the sun only determined every hour, at smaller timesteps or everytime it passes through a sky segment or ...?
If you don't have time for long explanations now, please run the calculation again first - on december 17th- and see whether you get the same results I get.
What I sometimes want to do with my models is rather checking the maximum insolation over a day than over a long period.
cheers