I have the following error when using the radiant ceilings from Ironbug.
** Severe ** InitConductionTransferFunctions: Found Material that is too thin and/or too highly conductive, material name = GENERIC PAINTED METAL
** ~~~ ** High conductivity Material layers are not well supported for internal source constructions, material conductivity = 45.000 [W/m-K]
** ~~~ ** Material thermal diffusivity = 1.427E-005 [m2/s]
** ~~~ ** Material with this thermal diffusivity should have thickness > 0.16028 [m]
** ~~~ ** Material may be too thin to be modeled well, thickness = 1.50000E-003 [m]
** ~~~ ** Material with this thermal diffusivity should have thickness > 3.00000E-003 [m]
I used the material setting you provided in many of the discussions in the forum but got this error and I actually cannot just increase the thickness of the metal.
Do you have other suggestions?
When I reduce my building to 1 storey. I have this error:
Program Version,EnergyPlus, Version 23.2.0-7636e6b3e9, YMD=2025.05.22 16:14,
** Severe ** [Branch][WaterLoopCooling Demand Branch 7][components][0] - Missing required property ‘component_name’.
** Severe ** [Branch][WaterLoopCooling Demand Branch 7][components][0] - Missing required property ‘component_object_type’.
** Fatal ** Errors occurred on processing input file. Preceding condition(s) cause termination.
So when I had 2 storeys I had the error I was describing in my last message and when I only kept 1 storey I got this error above.
It’s very hard to know where the problem is coming from.
Hi @fatmad you can create a new material with more realistic property values to replace this “GENERIC PAINTED METAL” layer.
That looks like you didn’t update the Ironbug ThermalZones when you switch 2-floor building to a 1-floor building. Please make sure your HVAC system is set up correctly based on your model.
Hi @MingboPeng,
thanks for having a look at the problem. I’m working on the same project.
We’re actually replacing the Generic Painted Metal layer with a layer that just meets the thermal diffusivity requirement from the log. This resulted in an unrealistic 16cm layer, but at least we can proceed with the debugging:
We’re getting this severe error:
** Severe ** CheckWarmupConvergence: Loads Initialization, Zone="GROSSRAUMBUERO_4AAA2F0D" did not converge after 25 warmup days.
** ~~~ ** ...Environment(SizingPeriod)="REGENSBURG-OBERHUB FEBRUARY .4% CONDNS DB=>MCWB"
** ~~~ ** ..Max Temp Comparison = 4.74E-002 vs Temperature Convergence Tolerance=0.40 - Pass Convergence
** ~~~ ** ..Min Temp Comparison = 0.00 vs Temperature Convergence Tolerance=0.40 - Pass Convergence
** ~~~ ** ..Max Heat Load Comparison = 0.1079 vs Loads Convergence Tolerance=4.00E-002 - Fail Convergence
** ~~~ ** ..Max Cool Load Comparison = 2.7071E-002 vs Loads Convergence Tolerance=4.00E-002 - Pass Convergence
Here’s what we tried so far:
We double checked all constructions and they seem to be correct.
We also applied Generic Constructions instead
We modelled a detailed HVAC and debugging purposes replaced it with a template, but the error remains.
We deactivated Natural Ventilation
We set all rooms to conditioned and applied the HVAC (in the real building we’re excluding some rooms. They’re not conditioned, have not setpoints and no HVAC applied to)
We increased the HeatingDesignCapacityPerFloorArea for our Heating Coil by factor 10 with no effect
Are there any ideas on your side?
If you’re fine with it, we’d be happy to equip you with the cleaned gh via DM.
Designing radiant systems that can consistently meet thermostat setpoints is a challenge, which I think many of us have come to accept is not feasible for all climates and building types. This seems to be part of your issue here but I have a suggestion to help test whether there are controls that might work well for your case here.
Can you try applying the same template Radiant Ceiling that you see in this shoe box example to your building?
You’ll see that the sample will simulate without severe errors and only a few unmet heating hours in winter mornings. The controls that get used in this latest template radiant system come from the OpenStudio-Standards gem and they have been informed by a lot of the latest research on radiant system controls that was done at LBNL.
If this system works, that will at least give us an OSM that we can try to replicate with Ironbug if you want to change more detailed properties of the system beyond what the template system supports.
thanks a lot for the advice! The template you shared works indeed! This is quite interesting, since we’ve tried just the DOASHVAC template before to check if the issue lies in the way we modelled our HVAC and it didn’t work out without the RadiantHVAC node.
However, the HVAC autosizing and control design seems to be quite off. With the template we get 800h of unmet heat and the cooling kicks in on winter evenings for some reason:
Also, instead of a boiler we’d like to use a heat pump, so now we need to figure out how to adjust our system to make it work. Unfortunatly just using our DetailedHVAC and then adding a RadiantHVAC node still produces convergence issues.
When using the “DOAS with low temperature radiant chiller with air source heat pump” template instead of the radiant chiller template you suggested we get those errors:
** Warning ** Construction RADIANT CEILING METAL CONSTRUCTION must have a temperature calculation that is between two layers
** ~~~ ** The temperature calculation after layer parameter has been set to one less than the number of layers.
and
** Warning ** GetElectricEIRChillerInput: Chiller:Electric:EIR="90.1-2019 WATERCOOLED CHILLER 0"
** ~~~ ** Energy input ratio as a function of temperature curve output is not equal to 1.0 (+ or - 10%) at reference conditions.
** ~~~ ** Curve output at reference conditions = 0.881
nothing severe, but still there seems to be an issue with creating the sql. RoomEnergyResults throws:
Solution exception:Length of values does not match that expected by the header analysis_period. 1424 != 1440
and EUI throws:
Solution exception:Failed to find the "Building Area" table in the .sql file.
If you are getting the simulation to finish with 800h unmet hours, the this is most likely not something you can blame on poor autosizing but rather your building represents one of those case where a radiant system is “not feasible for all climates and building types.” Because radiant systems use the thermal mass of whatever radiant surface they are heating/cooling, the rate at which they can change the level of heating or cooling they apply to a space is very limited. Radiant systems also need time to switch between cooling mode and heating mode, meaning they cannot respond to large changes in loads like the sun suddenly blasting through a fully-glazed facade.
Typically, radiant systems can only be used well in spaces with overall low (or very constant) internal loads that are always being kept at a constant temperature. Residences tend to be a good fit as they have low internal loads and (often) have small windows for the sake of privacy. Museums can also be a good fit since they also need small windows to avoid damaging artifacts and there’s a goal of keeping conditions very constant for the sake of preserving them. Maybe a set of closed offices with low people density and small windows could also work but a fully-glazed open office is almost never going to work just given how volatile the solar load and the people/equipment loads are.
So I would first recommend trying to re-design your system to minimize or eliminate the unmet hours with the template system. You can try things like making the windows smaller, adding shade, and increasing insulation. If that doesn’t work, you may have to accept that the type of program you are working with is not a good fit for a radiant system.
