Internal Temperature constant with no HVAC on


I’m trying to run a comparison model between Rhino and IES. I have no mechanical ventilation or infiltration, and a simple Ideal air HVAC system. I have the internal gains aligning but when I look at a single day I notice that the rooms temperature does not reduce, even though the external temperature is at least 5 degrees lower and there is no internal gains (and I have standard construction values). In IESVE the temperature does decrease.

Any suggestions on where to look? I looked at both a no-mass construction and a custom construction for the external construction subset but these don’t change the internal temperature during the hours that the HVAC is off.

Hi @Shanice,

There’s a lot of factors it could be driven by. My first bet would be U-values as you didn’t mention them, but I’m guessing that’s probably not the whole answer. To check the impact of constructions you could look at the external conduction gain for both and compare.

An approach I like is to report all possible gains/losses for a space - then you should be able to see whether they total a loss/gain as a sense check for changes in temperature.

The other factor could be your solar gains are relatively significant compared to your internal gains, so differences in thermal mass and solar absorption of constructions may be making an impact (with HB maybe absorbing more and releasing it at night).

That’s where I would start but may be more to it, worth triple checking your inputs, gains, etc. Seeing what happens at the end of the day would also be useful.

Hope that helps,

1 Like

Hi @charlie.brooker

Thanks for the reply. My U-values are Uwall - , U-roof - , U-floor - . They are the same in both programs but have very different external conduction gains which I cant seem to figure out why. I did do a total gain/loss sense check and that’s why I’m confused to why the Rhino model isn’t dropping temperature (even with the different conduction losses). See the graph, both IES and Rhino are negative losses (although im unsure why the window conduction would be negative during a summer day as it should be gaining during sunlight?)

If I look at the energy model example provided by Chris they have the same profile of no heat loss during the early hours of the morning.

full_building_energy_model (1).gh (429.5 KB)

Hi @Shanice,

Honestly I’m finding it a little tricky to follow without being able to dive in and interrogate results myself - I’d be happy to have a quick look at your script and IES model if you’d like (I’ve been doing my own tests recently to understand differences between HB and IES results so it’s useful learning for me). You’re welcome to send them to me in a private message if you’d prefer not to post them completely publicly.

If the U-values are the same in both it might be worth trying to copy the construction build ups as closely as possible between the software to reflect where the thermal mass and insulation lies within the construction. That may have some effect. Also checking the external solar reflectance / absorption properties in both might help align them.

From what you’ve posted it looks like there’s more loss at night from the Rhino model, so agree it’s strange to see the temperatures increase overnight rather than decrease.

The behaviour in the evening suggests that the heating set point in your IES model drops at 5pm, whereas it looks like it might stay the same in you Rhino model as it is during the day. Although worth noting that Chris has just changed how external conduction gain is being reported in HB to align with IES, perhaps you don’t have the latest version that includes that change?

Details of that here:

The energy model provided by Chris makes sense to me. I’m guessing that there’s a set back temperature of ~15C in that model, but I’ll have to check in GH when I’m at my laptop tomorrow.

Sorry this doesn’t really get to your problem, but hopefully the discussion helps the process of finding the difference.


1 Like

Hi @Shanice,

Thanks for linking me your script.

Some thoughts:

  • Side note - I just found out IES’ definition of U-values is HB’s definition of U-factors (HB defines U-values differently, not including the air film resistance, see here). I think that the way you’ve set up your model is probably correct between the two softwares, but thought you might find this useful to know as it was new to me.
  • I think it would be worth starting from a single zone model to make it easier to compare to start with. Multi zone models bring in adjacencies which add further complexity to the models and potential differences.
  • Ground floor U-values and adjacencies are a common source of differences between models. IES by default will have air as the adjacency for ground floors, whereas it looks like HB uses ground temperature (I know ground temperatures vary with depth and I’m not sure which it uses). It looks like the room you’re testing in detail is a ground floor room so this could have a significant impact.

I’m going to have to leave it there for now, but will hopefully pick this up again and add to this post.

1 Like


In addition to @charlie.brooker’s excellent investigation, one thing that sticks out to me is the differences in the HB conduction transfer versus its IES equivalent (although I’m not sure which construction elements the IES conduction term is measuring exactly).

Is the thermal mass of all the constructions equivalent? If the IES construction has little thermal mass relative to HB, but equal thermal resistance, you would expect the conduction transfer to be more sensitive to the outdoor temperature, result in more heat transfer overall, and higher diurnal peaks and valleys. This last one could start to explain why IES interior temperature ramps down during the evenings whereas HB doesn’t.

The EnergyPlus default for undefined ground temperature is a constant 18C, which is their rough assumption for ground surface temperature when accounting for heat loss from conditioned buildings in heating dominated climates. HB also uses this default - you can see the warning from EP in the .err report about this. I assume IES uses the average ambient air temperature as the ground temperature? Not sure how big of an impact this will have without knowing anything about the climate you’re simulating in, or the size of your building, but this could also be a contributor.

I second this. Reducing the complexity as much as possible will help. I would add that It’s hard to speculate without having any contextual information, since there’s so many possible degrees of freedom. I may have missed it, but can we get the city, and program type, and an image of the building?


Thanks for your insight, it certainly has been a eye opener on how similar but different the two programs run.

I have found that it is important to add the “Apply setpoint values”, otherwise the you can put in the same building information and get two different outputs (See attached model). (162.5 KB)

Setup One


Setup Two



That might be a bug, those two setups look equivalent to me, as I would expect the same default Office program filling in values you didn’t assign explicitly.

You can control this on your side a little better, though. Personally, I limit my reliance on defaults as much as possible, and try to as much parameters as explicitly as possible, as upstream as possible. This reduces the chance of conflicts between the default HB values and your manually assigned ones…

So in this case, if you use the HB Search Program component to set a program explicitly in the base_program_ or _program_ input, it makes all the program values consistent.



Also, was the difference in program templates what caused the difference in your internal temperatures? Did you get a chance to confirm if the constructions are the same?