Energy Plus also performs lighting simulations. I wanted to ask for your opinion about accuracy/validation of Energy Plus when it comes to lighting simulation.
I have read that Energy Plus has integrated DElight software (radiosity based lighting simulation software) and that measurements are within 10% error? Is this margin of error acceptable?
Depends on what you're trying to accomplish, but 10% error is perfectly acceptable for most architectural/lighting design simulation problems. But to simply say that the program is within 10% without stating assumptions regarding the various simulation parameters (which affect accuracy) is a gross misrepresentation (and, sadly, a common one with software manufacturer claims).
Our building is very simple (10’x15’ in the plan), and it has a 6'X6' window on a 10' X 11' wall. We expected sufficient
levels of illuminance for this room but we got low illuminance value of 270 lux at the workplane close to the window.
I am concerned is that either Energy Plus is not accurate enough or we have some error in our input.
What time of day? What sky condition? Hoe close is "close to the window"? Lots of questions...
How would you compare Radiance and Energy Plus (or DElight) when it comes to simulation of blinds?
We are new to Radiance and work already in Energy Plus, so we want to see if we can use Energy Plus for lighting simulation instead of Radiance.
Or Radiance is still the most accurate tool for lighting simulations?
Using a light-backwards ray tracing algorithm to solve light transport through blinds is a tough problem. Radiance includes a tool called mkillum that can do an initial raytrace at the blind interface and convert that result into a direct distribution for the window/blind system. This in turn is then used as a virtual light source to calculate the illuminance in the space. It leads to faster and more accurate results in Radiance. I cannot speak to Energy Plus' lighting calculation capabilities, but my experience with DOE-2 is that building energy simulation programs are severely limited in their ability to accurately estimate annual daylight and electric light performance. It is possible to precalculate the daylighting performance in Radiance and convert that to some sort of lighting schedule that can be plugged into the E+ model to at least get a single picture from E+ for the building's performance.
- Rob Guglielmetti
···
On Jun 5, 2008, at 11:34 AM, Svetlana Olbina wrote:
Hello Svetlana,
Naturally, asking on the Radiance mailing list, you're going to hear people say Radiance is the most accurate. In your case, I would say the 10% error is going to be overshadowed (no pun intended) by the variability of daylight. Specifically, it is very important to consider the weather data used as your basis in determining exterior daylight conditions. Even getting within 10% of measured illuminances requires the use of measured sky distributions, which are almost never available. So, I'd say this accuracy is based on the assumption that your sky matches the standard CIE model, which of course it won't if it's a real sky. Radiance would fare no better.
Annual simulation requires quick calculations, which is not really Radiance's strength. I know that Rob Hitchcock (the author of DElight) has put a lot of energy into his fenestration calculations, so if you aren't getting the output you expect, I'd take a closer look at the input as a first step.
Cheers,
-Greg
···
From: Svetlana Olbina <[email protected]>
Date: June 5, 2008 10:34:35 AM PDT
Hello, All,
We are doing research on blinds that we call “split controlled blinds”.
Basically, we divided a window in three sections in a vertical direction.
In each window section we use the same type of the blinds but the slats have different tilt angles in each section (for example, 0 degrees in the upper part (completely open) , 15 degrees in the middle part and then 90 degrees in the lower part (completely closed).
We plan to do both energy and lighting simulations for these blinds. For the energy simulations we use Energy Plus.
Energy Plus also performs lighting simulations. I wanted to ask for your opinion about accuracy/validation of Energy Plus when it comes to lighting simulation.
I have read that Energy Plus has integrated DElight software (radiosity based lighting simulation software) and that measurements are within 10% error? Is this margin of error acceptable?
Our building is very simple (10’x15’ in the plan), and it has a 6'X6' window on a 10' X 11' wall. We expected sufficient
levels of illuminance for this room but we got low illuminance value of 270 lux at the workplane close to the window.
I am concerned is that either Energy Plus is not accurate enough or we have some error in our input.
How would you compare Radiance and Energy Plus (or DElight) when it comes to simulation of blinds?
We are new to Radiance and work already in Energy Plus, so we want to see if we can use Energy Plus for lighting simulation instead of Radiance.
Or Radiance is still the most accurate tool for lighting simulations?
Thank you.
Svetlana
Hello Greg and Rob,
Thank you both for a quick response.
Rob Hitchcock's claims 10% error in his paper about DELight2 (DELIGHT2 DAYLIGHTING ANALYSIS IN ENERGY PLUS: INTEGRATION AND PRELIMINARY USER RESULTS).
Hitchcock stated that:
"Delight participated in validation exercise specifically focused on the ability to simulate
CFS performance, as characterized by measured BTDFs (Maamari, et al., 2005). The comparisons were based on measurements and corresponding simulations in a simple-geometry test box for combinations of CFS BTDFs and measured skies. The DElight2 results for a particular combination of a CIE overcast sky and a measured BTDF representing Serraglaze were shown."
This is all they mentioned about a sky model and the error. We used TMY 3 weather file downloaded from the Energy Plus website and did not add anything specifically related to the sky model in our input.
I wonder if this information from Hitchcock's paper warranties accuracy of the results.
We have not done experimental testing of our system and can rely only on the computer simulations at this moment....so we want to make sure we are on the right track when selecting a software for lighting simulations.
Thank you very much.
Svetlana
···
--- On Thu, 6/5/08, Greg Ward <[email protected]> wrote:
From: Greg Ward <[email protected]>
Subject: Re: [Radiance-general] lighting simulation of blinds in Energy Plus vs. Radiance
To: [email protected], "Radiance general discussion" <[email protected]>
Date: Thursday, June 5, 2008, 2:11 PM
Hello Svetlana,
Naturally, asking on the Radiance mailing list, you're going to hear
people say Radiance is the most accurate. In your case, I would say
the 10% error is going to be overshadowed (no pun intended) by the
variability of daylight. Specifically, it is very important to
consider the weather data used as your basis in determining exterior
daylight conditions. Even getting within 10% of measured
illuminances requires the use of measured sky distributions, which
are almost never available. So, I'd say this accuracy is based on
the assumption that your sky matches the standard CIE model, which of
course it won't if it's a real sky. Radiance would fare no better.
Annual simulation requires quick calculations, which is not really
Radiance's strength. I know that Rob Hitchcock (the author of
DElight) has put a lot of energy into his fenestration calculations,
so if you aren't getting the output you expect, I'd take a closer
look at the input as a first step.
Cheers,
-Greg
> From: Svetlana Olbina <[email protected]>
> Date: June 5, 2008 10:34:35 AM PDT
> Hello, All,
>
> We are doing research on blinds that we call “split controlled
> blinds”.
>
> Basically, we divided a window in three sections in a vertical
> direction.
>
> In each window section we use the same type of the blinds but the
> slats have different tilt angles in each section (for example, 0
> degrees in the upper part (completely open) , 15 degrees in the
> middle part and then 90 degrees in the lower part (completely closed).
>
> We plan to do both energy and lighting simulations for these
> blinds. For the energy simulations we use Energy Plus.
>
> Energy Plus also performs lighting simulations. I wanted to ask
> for your opinion about accuracy/validation of Energy Plus when it
> comes to lighting simulation.
>
> I have read that Energy Plus has integrated DElight software
> (radiosity based lighting simulation software) and that
> measurements are within 10% error? Is this margin of error acceptable?
>
> Our building is very simple (10’x15’ in the plan), and it has a
> 6'X6' window on a 10' X 11' wall. We expected sufficient
> levels of illuminance for this room but we got low illuminance
> value of 270 lux at the workplane close to the window.
>
> I am concerned is that either Energy Plus is not accurate enough or
> we have some error in our input.
>
> How would you compare Radiance and Energy Plus (or DElight) when it
> comes to simulation of blinds?
>
> We are new to Radiance and work already in Energy Plus, so we want
> to see if we can use Energy Plus for lighting simulation instead of
> Radiance.
>
> Or Radiance is still the most accurate tool for lighting simulations?
>
> Thank you.
>
> Svetlana