Negative Conduction from External Wall in Tropical Country

[kumudini1]

Hi

Im running simulations for a building model in Singapore, using the SingaporeIWEC.fwt weather files. Building is naturally ventilated. I have set it to naturally ventilated in ApacheSim as well.

And my external walls have an overall U value of 3.1649 W/m2K and medium weight thermal mass (185 kJ/m2K). The indoor air temperature of the building from simulation results closely follow the actual temperature observed in the building. However the conduction gain from external wall becomes negative during the daytime and with that when I calculate the external wall outside surface temperature based on the conductivity of the wall, the surface temperature is even lower than the outdoor dry bulb temperature.
eg. at noon, outdoor dry bulb temp = 32 deg Celsius. But the outside surface temp at that time is about 25 deg Celsius. But what is actually observed in site is about 45 deg celcius.

I understand there can are slight variations in simulation results due to the variation of actual weather condition and the weather files. However this deviation is way much off from the actual and I'm having problems in calibration and validation of the model.

i would be much thankful if some one could help me to solve this issue.

Thank you.

[Complex Potential]

Hiya

I suspect what you are seeing is down to the thermal lag in the wall itself. Often the temperature on the internal surface of the wall will be down to what was happening outside several hours before, sometimes even during the night which is why it acts to cool the space during the day.

Just my two cents.

[JohnLloyd]

You see this relatively often with IES and it's usually linked to periods of high solar gain. I believe what you're seeing is the thermal mass absorbing the heat; IES accounts for this within the calculations using the Conduction Gain variable.

[kumudini1]

Hi,
Thanks for the replies.

For IES, If i consider inner surface temperature (Ti) and the conduction gain (Q) from a specific wall in a given time from vista results, and if i use that 2 values to calculate the outside surface temperature (To) based on the wall thermal transmittance Value (U) and area (A), would it be an appropriate approach to obtain the outside surface temperature at that instant?

Thank you

[geshen516]

you can not calc the external temperature by using conduction heat gain and inner surface temperature.
This method is for steady heat transfer not transient.

[Tom_McGrath]

Hi

Could anybody elaborate on the answer given by geshen516, I'm particularly interested in this especially if you consider the heat conduction at an instant in time and any ignore effects of heat lag through the material.

I have a situation whereby I want to verify the conduction gain through a ceiling in IES and want to compare this against a hand calc if possible.

I'm considering either using the air temperature either side of the ceiling or the top and bottom surface temperatures (at a given time). Then using the steady state, one dimensional heat transfer equation to analysis and compare. Is this the right approach?

I had assumed that because the simulation can be quite large (up to 60min) this would be applicable?

[Complex Potential]

When running a dynamic thermal model the external surface to internal surface heat loss/gain is going to happen over the course of several hours depending upon your construction. For a particularly thin and lightweight contruction you may find that ignoring the heat lag effect results in similar results to the dynamic model but not with decent sized constructions.