Question-and-Answer Resource for the Building Energy Modeling Community
Get started with the Help page
Ask Your Question
0

Openstudio/Energyplus is not very accurate

asked 2025-11-11 09:47:25 -0600

Dutchie's avatar

updated 2025-11-25 10:30:48 -0600

Aaron Boranian's avatar

Just for fun I have build a digital model in Openstudio with FloorspaceJS of my house. After several simulations and tinkering with it a bit I have come to the realization that the simulation isn't very comparable to the effects I have noticed in the real world in my house.

For example: my living room is 485 ft2 (45m2). The floor, walls and ceiling are all about 8 inches (20cm) thick made of reinforced concrete (minus the windows of course) and yes I am aware that I live in a bunker. Now the results of several Openstudio/Energyplus simulations say that my living room can be heated up (air temperature) from 61 fahrenheit (16 celsius) to 68 (20 celsius) with a 2000w electric furnace in just 10 minutes (!) when the thermal mass of all that concrete is at 61 degrees.

Now I have just tried this in my living room with a electric 2000w heater and the model is no way near accurate. The model also suggests that when the room has been at 68 degrees for 8 hours and then the heating is turned off that the air temperature is back at 61 degrees in one hour. Which also is completely unrealistic.

Somehow it seems that the interaction with the concrete thermal mass is all wrong.

To be safe I made a new fresh and simple model. Just a square concrete box/room with an 2000w electric furnace. Same result. I don't understand. It seems very strange.

Does anyone have any idea about this or has had any experience with such a model?

Thank you.

edit retag flag offensive close merge delete

Comments

2

If you post a link to your model it might be easier to diagnose the inaccuracies. Some items that come to mind based on your description:

  • Are you accounting for infiltration?
  • You mentioned you modeled only your
    living room; are you accounting for
    air exchange and/or conduction
    between the living room and other
    spaces in your home?
  • Did you properly characterize the materials in your
    walls/ceiling/floor including windows and doors? What about other mass in
    the zone (e.g. furniture)
  • Are you accounting for internal gains?
  • Weather file
ericmartinpe's avatar ericmartinpe  ( 2025-11-11 12:26:59 -0600 )edit

Yes of course. I should have sent you the model. https://we.tl/t-8PvjXzR6xK

In the simple concrete box test (which I now have uploaded) I tried to make the model as simple as possible so there is no air infiltration, no other rooms, no internal mass besides the concrete.

Take a look at the 6th of february for example. The room is heated up in 6 minutes from 62 to 68 in 6 minutes!

Thank you for your interest. Much appreciated.

Dutchie's avatar Dutchie  ( 2025-11-11 16:59:04 -0600 )edit

You are right. I should have added the model. Here it is: https://we.tl/t-wFIt8WYFoV

As you can see in the file I have made a concrete box just to make sure there are not other influences other then the concrete thermal mass and some windows.

I really don't understand what is going wrong with Openstudio/Energyplus.

Dutchie's avatar Dutchie  ( 2025-11-12 07:10:00 -0600 )edit

Just curious, how did you made a weather file?

sandip_mep's avatar sandip_mep  ( 2025-11-20 22:13:19 -0600 )edit
add a comment

1 Answer

Sort by ยป oldest newest most voted
3

answered 2025-11-12 12:32:42 -0600

ericmartinpe's avatar

Here's one clue:

   ** Warning ** GetHTSurfaceData: Surfaces with interface to Ground found but no "Ground Temperatures" were input.
   **   ~~~   ** Found first in surface=FACE 0
   **   ~~~   ** Defaults, constant throughout the year of (18.0) will be used.

You need to specify ground temperatures for proper heat exchange with the floor and the ground, otherwise it assumes the ground is 18C/64F year round.

Also, your walls, floor, and ceiling are R~30, which is extremely well insulated (esp for walls and floors), so I'm not sure that's accurate.

Infiltration can play a significant role in thermal loads, so excluding that and assuming the space is tight with no air leaks, especially around the large window, is probably an inaccurate assumption.

edit flag offensive delete link more

Comments

Thank you for your interest in my question Eric.

I don't understand what the ground temperature has to do with the room heating up so extremely quickly. As you say.. the insulation is excellent so the ground is almost thermally disconnected from the room. So ground temperature wouldn't have an effect.. right?

Have you run the simulation Eric? Did you not find it odd that a room that size with such a cold thermal mass is heated up so quickly?

Dutchie's avatar Dutchie  ( 2025-11-12 19:11:08 -0600 )edit
2

The ground is a significant source of heat loss, which your model ignores. You've essentially modeled a completely air-tight, highly insulated box, which is not an accurate characterization of most homes.

Another unrealistic aspect is the electric furnace. The coil is 2 kW, but the fan runs continuously at 600 cfm, which is providing over 9 air changes per hour (ACH), at temperatures exceeding 35C/95F.

In short, I think the unexpected model results are more likely caused by an unrealistic characterization of the actual building/HVAC, rather than inaccuracies in the simulation engine.

ericmartinpe's avatar ericmartinpe  ( 2025-11-12 20:52:01 -0600 )edit
1

Great answer Eric!

antonszilasi's avatar antonszilasi  ( 2025-11-14 17:44:54 -0600 )edit

Thank you for your answer Eric. I am going to play with the model tomorrow. There might be something in your idea about fan cfm. I will also try the ground temperature to be sure and change the insulation.

ChatGPT and AIstudio are of the opinion that it has something to do with 1) wrong convection coefficients and/or 2) wrong HeatTransferAlgorithms and/or 3) energyplus 'fully mixed air assumption.

I will let you know.

Dutchie's avatar Dutchie  ( 2025-11-14 19:09:48 -0600 )edit

PS: Just did a experiment now. Took a temperature reading with a laser from my floor, walls and ceilling. All around 63f/17c. Room air temperature 63f/17c. No solar or other significant internal heat gain. No serious airflow to speak of. Turned on my 2.000watt electric fan heater and used a fan assisted matched temperature sensor which I know is very accurate. Result: the room air temperature heated up about 1c degree every 30 minutes. So taking the room from 63/17c air temperature to 68f/20c took about 1.5 hours. Not 6 minutes!

Dutchie's avatar Dutchie  ( 2025-11-14 19:10:06 -0600 )edit
see more comments

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Training Workshops

Question Tools

1 follower

subscribe to rss feed

Stats

Asked: 2025-11-11 09:37:48 -0600

Seen: 239 times

Last updated: Nov 12

About | Help | Privacy | Legal | Contact
Site design and logo: Copyright © 2015 Big Ladder Software LLC. All rights reserved.
The Unmet Hours and Big Ladder names and logos are trademarks of Big Ladder Software LLC.
User contributions licensed under the Creative Commons Attribution Share Alike 3.0 License.
Powered by