I have a questing regarding the Solar Exposure Calculator and the data behind the thematic display style.
As a generic scenario, we have a concept and we would like to calculate the solar exposure for say some typical apartments.
As an example lets say it is this simple box with a few windows/openings.
We are interested in the amount of solar exposure on the floor on a single day of the year between 9am-3pm.
So running the Solar Exposure Calculator gives us a graphic answer like this:
Which is ok.
With a bit of adjustment to the setting of the displaystyle and turning off the walls and roof we can focus on the floors.
As you can see we are interested in areas that have solar exposure greater than 2hrs so we can just use 2 colours to visually see the impact.
Now for the question:
Is there a way of accessing the data behind this display? Is it stored somewhere in the file or element?
Purpose of this is 2 points.
I could export multiple models every 15min to get the information needed in point 2 if we could measure the area in point 1 but it would be nice not to have to export all those models if we could analyse the data.
Any help would be appreciated
Unfortunately you cant simply extract boundaries to quickly get the Surface Area of interest.
You might find some of the info in the Help of interest.
Instead of outputting the results to a design file you can create a csv file.
The csv file will contain information that pertains to each facet in the mesh that is created the useful part will be the coordinates for the facet, also it is the total so you will have to run multiple studies to show the change in the 15 min inrements desired.
From within excel you can filter facets below exposure threshold. To determine area it will take a good amount of work will have to brush up vectors and cross product, will have to create a formula to calculate area of each facet and then create a total. Again this will get very complex as the facets created can be triangles or squares so its not just one formula that you will have to create.
Answer Verified By: Wayne Dickerson
So some success in getting information out, which raises some more questions regarding the CSV file data.
Looking at the results for just a single output for 15min on a single day etc.
I can set the thematic display to show a nice graphic of what area gets the full 15min of sun during that time.
What I see in the CSV is only a clear 900 "ExposedSeconds" or 900 "ShadowedSeconds", I would have expected a slight change on the points?
When you look at the display style in a finer breakdown you can see a change across the edge.
Is that a setting in the display style or an issue with my CSV export.
(they were done with the same settings CSV first then Design File without changing settings)
Reason I ask is I managed to get the CSV process by GC and produce some polygons so I could measure the areas of direct sun exposure, which is great but the areas don't match-up from the graphic display vs GC numbers. Hence the reason at looking more closely at the CSV file.
As you can see the shadow polygons area slightly different vs the rendered view, I will keep checking the GC to makes sure the data is correct.
Attached CSV for reference.
I worked it out. I had set the increment to 15min hence the 900 default.
Setting to 1min gives a more refined result in the CSV.
Will do an update in GC and see the results.
Good to hear if possible could you share the final results and files for review. Should have thought of GC myself but not familiar with all the capabilities.
Just an update my solar experiment.
I attempted to push things as far as I could with some lessons along the way.
The outcome was to try and measure the area of direct sunlight on a surface (in this case the floor slab) that received 15min of sun in the period 9:00-9:15.
As I mentioned above I did a manual measurement of the visual display in the solar output design file as a guide, then I exported the data to CSV to process in GC.
This worked well but the data was not detailed enough and the areas in GC were much higher than measured manually. So I looked at reducing the sample size to get a better result. Which is where it started to hurt GC, at 0.1 metre resolution it produced around 450,000 rows in the file and I did attempt to run this through GC but after about 8hrs I decided there should be a better way.
So I wrote some VBA in excel to strip out the relevant data I needed from the 450,000 rows, filtered by element id and rows where solar exposure is >= 900 sec. This was very effective and in about 10-15sec it reduced the needed rows down to under 2500 for a single element id.
Putting this back into GC gave a nice set of polygons and a resulting area. which is close to the manual measured area.
This is the manual area I measured. Total around 10.82sqm
This is the results in GC from the CSV/Excel data. Total area 11.65sqm
To try and check the data in GC I went back to excel and wrote some more VBA to calculate areas of polygons and did the process in VBA. It turns out it is much more efficient to do it in Excel. I was able to do 3 floors in under 30sec
Added some buttons to do the work of refining the data to the areas needed and then calculated the areas.
As you can see the areas in Excel match the Areas in GC so I call that a success.
You could even use spaces from OpenBuildings as elements and then have excel matchup the IDs and you could have a nice schedule by room etc
It did make me think of trying to refine the data even more but I found there is a limit in the Solar Exposure tool of 0.05M or 50mm.
Running at this didn't really refine the results that much more and produced around 472000 rows. I tried to trick the system by scaling my model up by 10x to see if i could get more data but not really. Probably best as excel only really has 1M rows to use.
So as a summary there is a limit to how far you can push it, but it gives some very fast feedback even with a very basic model compared to say using the full energy simulator that requires a bit more setup to get results. (which may provided a much finer result, will have to try that one out next). Great for early design work.
Could I ask one last question if anyone knows how the visual display is determined? The reason is it looks more refined than the CSV data you can get, just wondering if they are both using the same results but the visual thematic display is interpreting it in another way.
If anyone wants to test out the process I have attached the simple model you can run the same tests on.
What a tour de force! Thanks for describing your solution to a complex problem.
Regards, Jon Summers LA Solutions