Discrepancy of the results of energy use for electric lighting

Replies to This Discussion

Permalink Reply by Abraham Yezioro on May 4, 2015 at 12:08pm

Ding,

Your findings are not a big surprise.

What i recommend you is to define in parallel the different parameters of the lighting for both E+ and Daylight simulations. When you get to the point that ALL parameters are equal you probably will get similar results. At this point, without checking your file thoroughly i can bet that there are basic differences and hence the differences in the results. Among the different parametres i would check are: Control, Lighting levels, Schedules, etc.

I myself will be glad this comparison between both simulation types. Please upload the file when you get there. I suppose that for this task Mostapha/Chris can help a bit. I think an example file like this can be very useful.

Good luck,

-A.

Permalink Reply by Ding on May 4, 2015 at 1:40pm

Hi Abraham,

Thank you.

What I want to add is that the lighting schedule which used by the E+ simulation was generated by the daylight simulation (namely "Honeybee_Read Annual Result I" component). This means both E+ and Daylight simulations share the same lighting schedule. So I assume their results should be more or less similar or not that different, even though there might be some difference in their calculation methods. 

But, having tried many times, the result generated by the "Honeybee_Daysim Electrical Lighting Use" is usually much less than that generated by the "Honeybee_Read EP Result" (e.g. 160.5 kwh vs 369.425 kwh as in the current parameter).

In addition, in "Honeybee_Lighting control Recipe", default parameters (lightingPower = 250, lightingSetpoint = 300, ballastLossFactor = 20, standbyPower = 3, delayTime = 5) were used in the current case except "control type" - [5] Always on during active occupancy hours with auto dimming.

In short, both E+ and Daylight simulations have already shared the same lighting schedule. As for Control type and Lighting Setpoint, they are only required in daylight simulation.

Best,

Ding

Permalink Reply by Abraham Yezioro on May 4, 2015 at 10:32pm

Hi Ding,

Both simulations may share schedules (light) but they definitely don't share daylight considerations. While the Radiance does it, the E+ doesn't at ALL. If you check the IDF created you'll see that there is no Daylighting class. So i assume that in your case the light is on all the time as specified by the schedule. That's probably the reason you get such a difference in the results.

Please see this discussion to get updated about the status of the daylighting in HB+E+:

http://www.grasshopper3d.com/group/ladybug/forum/topics/hb-daylight...

Right now it is not working, so you really can't rely in the E+ result so far.

-A.

Permalink Reply by Ding on May 5, 2015 at 5:32am

Hi Abraham,

Thank you for the link. Actually I had already been aware of it. This is why I am trying to run daylight simulation first then pass the generated schedule to the E+ simulation for obtaining energy use of electric lighting dimming system.

But I found a major mistake in my previous test file, that is, the "lighting Power" used in the daylight simulation and the E+ simulation are NOT equal by my carelessness. The former was 250w (default value in "Honeybee_Lighting control Recipe") while the later was 11.8403571*50=592.018w (lightingDensityPerArea*Areaoffloor=lightingpower). 

Having corrected this mistake (by using 592.018w for both of them), the results of electric lighting energy obtained from daylight and E+ simulations are much closer with each other and more reasonable. In order to see how close they are, I made 3 tests using 3 different weather files (New York, Amsterdam, Guangzhou). The results are listed as follows respectively:

New York (3.4% difference)

Amsterdam (2.7% difference)

Guangzhou (3% difference)

It seems that the difference (around 3%) is acceptable?

You are right. Having checked the IDF file created, as you mentioned above, there is no data in the Daylighting class (since the daylight control are not implemented in the E+ simulation). But, the light is "fully/partially" on according to the "fractions" in the generated lighting schedule list, which already takes in account the dimming system by the setting in "Honeybee_Lighting control Recipe" in daylight simulation.

Then, what the E+ simulation did in terms of calculating electric lighting energy use is just as the following formulation: lightingDensityPerArea * Areaoffloor * sum of lighting schedule list (generated by daylight simulation in this case) = electric lighting energy use, which has been verified by the comparison of results obtained from E+ simulation and calculated by the above formulation. (The results are exactly the same.) Please see the updated gh file attached. 

Therefore, I assume that the E+ results generated in this way (using the lighting schedule generated by the daylight simulation) are still reliable, although there exist some tiny difference (around 3%) for some unknown reason (Maybe Daysim does not follow the same formulation mentioned above? Can you imagine the possible reason?).

In addition, 

- I am not clear with what you mentioned as "they definitely don't share daylight considerations". What "daylight considerations" here refer to? (sky condition / sky file? but in the annual Daylight Simulation, only weather file is required instead of sky file)

- A basic question: In E+ simulation, has the waste heat generated by the electric lighting been considered as "internal heat gain" (showed in the above image)? If so, can I say that "energy use for electrical light" is fully converted to "the waste heat" + "Illuminance in lux".

Namely, energy use for electrical light = the waste heat + Illuminance in lux (lux can be converted to electric power in watts)?

Many thanks!

Best,

Ding

Attachments:

• text8.gh, 597 KB

Permalink Reply by Abraham Yezioro on May 5, 2015 at 9:13am

Hi Ding,

Your explanation sounds convincing and i feel more comfortable with it.

I checked the schedule file and i see that effectively, you have there the fraction of the light used for each hour. So that compensates somehow the lack of the Daylight class in E+. A difference of 3-4% is bearable. I checked also with my local EPW and i get the same range of difference.

No doubt the daylight simulation with Radiance/Daysim is more reliable than the E+, but just for you to know (maybe you do), E+ allowsup to 2 sensors in each thermal zone. It will be interesting to "downgrade" the daylight simulation to just one or two grid points (locates strategically in the space), instead the whole grid, and see how the results change. For calculation time in your model it won't make a difference, but if you have a more complicated model i will probably will select carefully the proper sensor location.

I'm not sure what do you call Heat Waste, but light in E+ also represent a source of heat (internal heat gain, such as the equipment) that will affect the Heating/Cooling loads.

Good luck,

-A.

Permalink Reply by Abraham Yezioro on May 5, 2015 at 9:55am

I'm not focused today.

I see that you picked just a couple of points for the anuual daylight simulation. Sorry.

-A.

Permalink Reply by Ding on May 5, 2015 at 5:42am

Btw, I assume some other people also use and tend to trust the similar procedure and the E+ results (setting up an E+ simulation based on daylight simulations.)

For instance:

http://www.grasshopper3d.com/group/ladybug/forum/topics/energyplus-...