[Archived phorum post from squ1.com v2 by Rehan]

Hi Oliver
its interesting to know that yesterday i was thinking about the same because i just did model the trees in my BIG MODEL as 2d and then the interzonal adj and other calculations didnt take too long time (although a bit long and the computer i have is the fastest i can have its p4 with 1024 ram and i am a bit afraid to overlock it for processor speed ) well what i want to model is that the trees be more tranparent in winter and less transparent in summer to give a bit closer effect of what can happen in real situation on site analysis features. I am on the same track of questioning i ended up in a bit mess last night while trying to do this if someone knows pl do help me
Rehan

[Archived phorum post from squ1.com v2 by Andrew]

Hi Olivier and Rehan,

I've just run a quick test assigning a schedule to both an appliance objects (to generate a burst of internal gain) and to the air change rate of a zone to simulate the opening and closing of windows. Both seemed fine - I could see a clealy defined temperature increase when the appliance was activated and could see the internal environment temperature move closer to outside temperature when the zone was activated (I used a maximum air change rate of 10ac/h to clearly see the effect). You might want to send over the model with a quick explanation of what you are trying to do and what you can't see for me to investigate.

Regarding the changing overshadowing from trees, when ECOTECT does a thermal calculation it caches the shading mask from surrounding objects. Thus to simulate variable shading of deciduous trees you need to do the following steps:

Step 1:
- Assign the trees the lowest transparency (ie: the maximum opacity) value that you want to use.
- Select one of the windows they will overshadow and click the Calculate » Sun-Path Diagram menu item.
- Calculate the Calculate Shading toolbar button in the Sun-Path Diagram dialog and generate an Overshadowing Percentage display.
- When complete, choose the Display » Display Shading Coefficients menu item and read off the average shading coefficient for the month when the trees will be at their most opaque.

Step 2:
- Assign the trees the highest transparency (ie: the minimum opacity) value that you want to use, and then do the same as Step 1 but read off the month at which the trees will be the least opaque.

Step 3:
You now have the variation in average overshadowing resulting from the trees - in the example I used, it varied from 21% to 71%. As the window itself already has a shading coefficient (SC) due to the glass and frame of 0.74, we need to moderate this value based on the additional SC of the trees.

- Thus, the minimum SC will be 0.74x0.21 = 0.1554.
- The maximum SC will be 0.74x0.71 = 0.5254.

Assuming that the transparency of the trees varies with season, you can take a relatively simple approach and use a schedule to vary the SC of the window by creating one window material with the maximum SC value and another with the minimum. If you assign one as the primary material of the window and another as the alternate, you can then use a schedule to vary between the two materials over the months of the year. You can do this with a simple step, by changing the between one 100% daily profile in summer and a 0% daily profile in winter, or you could even set up 12 different daily profiles for each month of the year, gradually varying between the two.

In the next release of ECOTECT you will be able to use scripts to modify schedules and shading masks interactively during your calculations, however in the interim this will give you pretty close to the same result.

Hope that helps...

Regards,
Andrew

Dr. Andrew Marsh
Square One research Pty Ltd

[Archived phorum post from squ1.com v2 by Rehan]

Hi Oliver and Andrew
First of all thanks alot for helping me out with stuff i cant do my self at all.
I noticed that out as well that calculation time is really because of the trees so now for the model i made the trees as a 2d object and it works good witht the calculation time. About scheduling the trees i am working it out and hopefuly i will manage it:
What i have done now for this model is that i placed the analysis grid at the area where the buildings are and then ran cummulative insolation.
when the insolation completes what i expect is that the area covered by buildings should be showing no solar radiation at all thus giving a blue colour corresponding to 0 Wh but its not the case in the inner sides of the buildings and there are no windows at all and i have run interzonal adj to see if there is some discrepency in model but couldnt find any. I will send this model with the analysis grid to u people.
my second question is about Wh.
see i make a simple calculation that for freiburg avg solar radiation is 1000 kWh/m2 per year so for per day it would be 1000/365 = 2.73 kWh/m2 as an average daily for 1m2 now when i multiply it for one m2 it will give me 2.73kWh or 2739 Wh but when there is an analyis grid or even if i do surface subdivision i get a value like 3000 Wh or something like this which can be acceptable if it is per m2 but if the grid dimensions are lets say 5m2 and i get an answer in Wh it should give me 2.73kWh/m2*5m2 = 13.698 kWh or 13698 Wh. I need a comment on these two issues from you or anyone who knows about it what it corresponds to.
best regards
Rehan

[Archived phorum post from squ1.com v2 by craines]

Andrew has said to let you know Olivier that you are absolutely spot on with the grid point versus surface plane for Rehan's Wh questions.

As you mention it would be more appropriate (if you are after Wh relative to area/surface etc.) to use the Cummulative Insol. not with the grid but using objects in the model (on thermal zones). Usually the best thing for this is to setup one object and then use the surface subdivision function (Modify Menu » Surface Subdivision » Rectangular Tiles) to get your objects to calc from.

With regard to the first issue you had Rehan, the colour of each grid cell is an average of all four vertices that make up the square. Thus if all four vertices are inside the building then you will have blue squares of value 0Wh, but if one or more points are outside the average will be greater.
One way to investigate this further is to turn off the Shade Grid Squares option, and turn on the Show Node Values instead. Then if you zoom into your model you will see that all the nodes that are inside do get 0Wh.
Becareful though because some of the nodes seem to line up exactly with the geometry of your buildings -- if a point exists exactly on some model geometry its results will be innacurate, as ECOTECT (being a dumb computer program) will sometimes think the point is outside, and other times think it is inside, so the value will not be useful for you.

To avoid all this hassle it is best to do a couple of things when setting up your Analysis Grid:
1. Always make sure you leave a small gap between any Analysis Grid point and model geometry, around 50mm is good. The aim is not to have any points line up exactly with any objects, as this will cause confusion for ECOTECT As a side, this is the reason why, when you use the automatic Fit Grid options in ECOTECT that it always offsets the grid inwards slightly.
2. If you don't want particular areas of your grid to show values, or you need a grid shape other than square, use the SHow / Hide Grid Nodes function (click Select Grid Nodes first). This way you will not only remove any nodes that you aren't interested in, the calculation will be quicker because ECOTECT won't calc. values for nodes it can't see, and you also won't see the confusing averaging that Rehan mentioned.

Small sample screen shot from Rehan's model
Regards,
Caroline

Caroline Raines
Square One research Pty Ltd

Hi,

I am tryting to model blinds which are shut from 3pm to 7pm from December to February. I have copied the window material and modified the SHGC from 0.53 to 0.2 for the window with blinds and the transparency from 0.74 to 0.2.

I have run the monthly loads report and the reults show that the cooling load decreases (as expected) but the heating load increases. I would expect some increase in heating load if the blinds were shut all year round but in the cooler months i would expect the heating load to remain the same seeing as the blinds are not drawn through this period (i am in Australia so that why blinds are drawn in Dec - Feb)

I have run the calculate shading tool for a point i set up on the floor which will be affected by the blinds to check what was happening but there is no change in the results with or without the blinds.

I am not sure what I am doing wrong as I believe I have set up the schedule and the window data correctly.

Any help would be greatly appreciated.

Andrew

Hi Oliver,

Thank you for your help. I have some aditional questions.

Do you mean that the primary material should be the non shaded window? As I only want shading from 3pm to 7pm.

Does the IZA need to be run every time there is a change to the schedule?

What should the Time on and Time off settings be under Activation (0.00 and 24.00?)

Regards,
Andrew

Hello,

I opened the file you sent as an example, Olivier, but I could not see a difference between the window's primary material and it's alternate material. Is that normal? I can't find where to change an object's shading coefficient...

Regards,

Alice

Ok, never mind, I found it!

Hello Andrew,

I've been reading your post over and over again and I still don't understand: how can you just multiply the window's SC by the SC caused by the overshadowing of the trees?
This basically means that if the window is very shaded, its SC will be very big, and it will let a lot of energy in. Isn't that wrong?
Shouldn't you multiply the window's SC by 1-the shading coefficients caused by the trees?

And another question I had about your post is: when you calculate the shading coefficients of the window without any trees, they're not equal to 0, which means that the shading coefficients you get with the trees are not entirely caused by them! Shouldn't you take this into acount when assigning a SC value to your window? Cause if not, won't it suffer the overshadowing of other things (building itself, other buildings, etc...) twice?

I hope I am being understandable... And hope you will answer me soon because I have really spent a lot of time trying to figure this out!

(what I want to do is calculate the g ("facteur solaire") of a window with its blinds, and I thought I could use your method for this..)

Thank you very much,

Alice

If your window has an SC of say 0.82, and the tree applies another effective SC of 0.6, then the overall SC will simply be (0.82 x 0.6) = 0.492. Thus, only 60% of the available light/radiation actually reached the window, and of that 60%, only 82% actually got through (hence: 0.82 x 0.6).

Higher shading coefficients actually mean more light/radiation passing through. For some standard SC value, see the following WIKI page:

http://squ1.org/wiki/Shading_Coefficient

Thus, you are absolutely right. The SC in the tables produced by ECOTECT are in error. There is supposed to be 7 columns, 3 for shading and 3 for shading coefficients. It seems that a bug has been included in the June 2007 update that omits half the columns. The real SC is, as you suggest, 1 - the displayed value.