Confused simulated air temperature

Post here for very basic questions and request for help...
But: BEFORE you post, read the older posts if your question hasn't been answered before.
Please be specific. Q like "I don't understand ENVI-met" will most likely be ignored.
Long
Posts: 10
Joined: Tue Jan 28, 2020 10:44 am

Confused simulated air temperature

Postby Long » Thu Mar 19, 2020 4:05 pm

Hello Envi-met team,

I did a research about the Urban Heat Island (UHI) effect on the thermal performance of a building located in the urban area in Wuhan, China. I simulated one extream hot day (August 09, with one day buffering time) in summer using Envi-met v4.4.4.

I found that during the night, the air temperature at the hight of 3 m of receptor 1 around the building is lower than the input temperature (from EPW file), as shown in attachment figure. It seems to be contradictory to the knowledge that the larger UHI effect is obsearved during the night than daytime (meaning the simulated temperature should be higher than EPW and the difference should be even larger than the daytime). Is this normal?

In case you need, I put my simulation files in the attachment. Thanks a lot for your kind response.

Best regards,

Long
Attachments
Diapositive1.JPG
Diapositive1.JPG (12.63 KiB) Viewed 41 times
Envi-met files.zip
(828.78 KiB) Downloaded 2 times

Tim
Posts: 1219
Joined: Mon May 15, 2017 3:46 pm

Re: Confused simulated air temperature

Postby Tim » Fri Mar 20, 2020 7:51 am

Hi Long,

That can have multiple reasons, maybe it is a combination of them. These two could help you with the simulation:
- The materials did not yet store enough heat which can be released during nighttime. Even longer simulations would thus be needed.
- The radiative fluxes are not calculated accurately enough to measure the effects causing the UHI. You would need to activate IVS to receive a better result. However, we know that it takes much longer to run a simulation with IVS and thus programmed a much faster algorithm which takes into account the surrounding surfaces. This will be an alternative for the very accurate but rather slow IVS algorithm. It will be available with the upcoming release.

Best regards,
Tim

Long
Posts: 10
Joined: Tue Jan 28, 2020 10:44 am

Re: Confused simulated air temperature

Postby Long » Mon Mar 23, 2020 11:04 am

Hello Tim,

Thanks for your suggestions.

I tried to prolong the initiallisation time from 1 day to 3 days (4 days in total simulation), but I found ths simulation crashed on the second initialisation day, as shown in the figure below. The timestep I set is 2,2,1.

This crash sometimes also happens when I simulated other days (meaning for some days it crashes, some not).

I tied to read the PANIC file but they are not readable (some strange code).

Do you have a suggestion to solve this problem? Increase the border cells will help?

Best regards,
Long
Attachments
1.png
crash information

Tim
Posts: 1219
Joined: Mon May 15, 2017 3:46 pm

Re: Confused simulated air temperature

Postby Tim » Mon Mar 23, 2020 12:23 pm

Hi Long,

PANIC files can be visualized in Leonardo.
Other days / longer simulation periods may include strange meteorogical conditions from the EPW file. You could check that in the ForcingManager. Especially fast and huge wind direction and speed changes are a problem.
I would definitely not recommend to increase the timesteps, rather decrease them to get a more stable simulation.
You could edit the MinFlowSteps for FullForcing in the ENVIguide to 10 to get a more stable run.

Best regards,
Tim

Long
Posts: 10
Joined: Tue Jan 28, 2020 10:44 am

Re: Confused simulated air temperature

Postby Long » Mon Mar 23, 2020 2:31 pm

Hello Tim,
Thanks for your reply. I will try your suggestion.

Here is another question: when I use ENVIguide to generate the SIMX file, I can choose PROJECT FEATURES in a window containing various features such as PLANTS, OUTPUTS, TIMESTEPS...

I notice that if I click the PLANTS model, it shows [use the tree calendar] is defalutly chosen. But when I simulated some months such as Feb, June or Nov, I didn't click click this PLANTS model, so did the simulations defaultly consider the tree calendar (different LAD)?

Long

Tim
Posts: 1219
Joined: Mon May 15, 2017 3:46 pm

Re: Confused simulated air temperature

Postby Tim » Mon Mar 23, 2020 2:50 pm

Hi Long,

Yes, Tree calendar is activated per default. You could disable it in the ENVIguide if you wanted to.

Best regards,
Tim

Long
Posts: 10
Joined: Tue Jan 28, 2020 10:44 am

Re: Confused simulated air temperature

Postby Long » Mon Mar 30, 2020 10:35 pm

Hello Tim,

Thanks for your answer. As abovementioned, I increased the initailization period (3 days) but the simulation crushed.

In fact I tried to decrease the MinFlowSteps in full forcing, but it didn't work. Finally I increased the border cells from 3 to 8, and it worked.
Does this solution make sense to you?

Best regards,
Long

Tim
Posts: 1219
Joined: Mon May 15, 2017 3:46 pm

Re: Confused simulated air temperature

Postby Tim » Tue Mar 31, 2020 6:41 am

Hi Long,

Yes, of course, problems with wind direction changes are especially caused by buildings near the border. Increasing the open space at the borders is always a good idea to increase stability.

Best regards,
Tim


Return to “Beginner's Corner”

Who is online

Users browsing this forum: No registered users and 0 guests