For air-conditioned buildings, the Predicted Mean Vote (PMV) model calculates a comfort score ranging from -3 (cold) to 0 (neutral) to 3 (hot).
For naturally ventilated buildings, the adaptive thermal comfort model (ATC) model calculates an indoor design temperature, around which a range is defined as comfortable for 80% (or 90%) of people.
Is there a way to compare these two different comfort metrics?
For example, if I simulate a space to be air-conditioned and determine a PMV of 1.5 at a given time, then I simulate the same space to be naturally ventilated and determine the indoor temperature to be 4K over the design temperature - in which scenario is the occupant more comfortable?
Asked another way: when using the ATC model, can I use the difference between the indoor temperature and the design temperature to determine a comfort score based on the same 7 point scale as the PMV model, in order to enable a direct comparison?
Kind regards,
Max
Unfortunately, a direct translation is not possible. What you could do is translate the PMV to a PPD value and then compare the PPD value to where your measured temperature lies on the ATC chart - in the 80% band or the 90% band. Again, there is no way to be more precise with the satisfaction value - you cannot say it is 85%, just that it lies between 80 and 90%. This is the interpretation from the standards, PMV and ATC are not equivalent methods for determining comfort.
Now, here I am going to suggest a little round-about way which is not something well accepted or validated. You use Griffiths' equation:
Tc = Tair - TSV/m
m is suggested to be 0.5 (Humphreys, M. A., H. B. Rijal, and J. F. Nicol. Building and environment 63 (2013): 40-55.)
Tc is the comfort temperature. In this situation, you can calculate it from the ATC chart. Tair you measure and thus you can calculate a TSV to compare with PMV.
Now this is a convoluted use of Griffiths' equation which is typically used to calculate Tc from survey votes. But it is something you can try.
You should use logistic regression for these variable and you cannot use linear regression because ATC is an ordinal variable and temp is scale.
Dear Max,
We have used different survey's scales to investigate the accuracy of thermal comfort votes.
Maybe the following papers would be useful for you.
Hope you find your answer in them,
"A satisfaction-range approach for achieving thermal comfort level in a shared office"
https://www.sciencedirect.com/science/article/pii/S0360132318303512
"Investigating Occupancy-Driven Air-Conditioning Control Based on Thermal Comfort Level"
https://ascelibrary.org/doi/abs/10.1061/%28ASCE%29AE.1943-5568.0000295
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude