Skip to main content
Open Access Publications from the University of California

Modeling thermal comfort with radiant floors and ceilings


The surface temperatures of radiant floor and ceiling systems should depend on the ambient air temperature, yet the surface temperature limits specified by current standards do not vary with air temperature. In addition, the limits for ceiling temperature are specified in terms of radiant temperature asymmetry, which is difficult to convert into surface temperatures. This paper provides graphs that allow designers to directly determine, for a representative room geometry, the acceptable range of floor- and ceiling surface temperatures as a function of air temperatures.

The graphs were generated using the Berkeley Thermal Comfort Model (BCM). Acceptable and optimal floor or ceiling temperatures were found for a range of air temperatures for normal office work activity level (1.2 met). Acceptability was defined as the absence of whole-body discomfort. Depending on the air temperature, the acceptable floor temperature range is 15-40ºC, wider than that specified in ASHRAE Standard 55 and ISO 7730 (19-29°C). The upper limit of 40ºC is based on avoiding discomfort through skin contact, supported by several studies showing that people are comfortable with floor temperatures near 40ºC. The 15ºC lower limit was chosen to avoid local foot discomfort as reported in a laboratory study. The acceptable ceiling temperature range is 10-50ºC, also wider than in Standard 55 and ISO 7730 (radiant asymmetry <5ºC for a warm ceiling, and <14ºC for a cool ceiling). The maximum temperature of 50ºC was chosen as a reasonable water temperature available from heat reclamation, and it was felt that temperatures below 10ºC were unlikely to be used in any climate because of the risk of moisture condensation on surfaces. The model simulation results were compared with published laboratory experiments on radiant floors and ceilings. The results are in generally good agreement, given a number of uncertainties in reproducing the laboratory conditions and matching a variety of different comfort voting scales.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View