I'm no statistician, but I've always been dubious about that calculation.

You are taking the mean of two means, which seems to be getting farther

and farther from actual data.

It also seems to punish some of the climates that are best suited for

natural ventilation, since having cool nights significantly drops the

acceptability limits. I assume they are pushing you to take advantage of

night pre-cooling and thermal mass, but I find it curious that there are

no allowable hours outside of the range. All of the pre-cooling in the

world isn't going to help you keep the temperatures down during that one

string of 90 degree days in the tmy2 file...

Nathan Miller

John Ross wrote;

Using a mean monthly temperature ties the hands of innovative engineers

in opportune climates.

In Northern Virginia I use an opening roof skylight to dump excessive

solar gain in the late afternoon.

This operational variation is only practiced the last week of September

to the second week of October

depending depending on the particular year's weather

It would be appropriate to use mean monthly temperature for calculating

long term Geothermal field

effects on the other hand.......

Mean monthly outdoor temperature was incorporated in the ASHRAE Standard 55-2004, because this parameter can be found easily from readily available climatic data, such as published by www.ncdc.noaa.gov

In the ASHRAE RP-884 project, which formed the basis of the adaptive thermal comfort in ASHRAE Standard 55, the outdoor temperature index used was the mean effective temperature (ET*). ET* combines temperature and humidity into a single index and for RP-884 was calculated based on measured data. It should be noted that RP-884 involved nearly 21,000 sets of raw data compiled from 160 different office buildings located over a broad spectrum of climate zones spread across 4 continents. For more information regarding the relationship between thermal comfort and the environmental index- ET*, please refer to the Thermal Comfort chapter in ASHRAE Handbook of fundamentals.

As mentioned earlier, the mean monthly outdoor temperature was introduced in the ASHRAE Standard 55, because it is easy to use and can be estimated quickly. The regressions (between the indoor operative temperature and the outdoor temperature index) that were finally incorporated in Figure 5.3 in the ASHRAE Standard 55 were recalculated based on mean monthly outdoor temperature (original index being ET*).

It is not necessary, that everyone agrees with the adaptive thermal comfort envelope yielded by using the mean monthly outdoor temperature and figure 5.3 of ASHRAE Standard 55-2004. According to one of the voting members of ASHRAE Standard 55 committee, the method was incorporated in the Standard because the majority of the members voted in favor of this method. And as you are all familiar with the public comment procedure of ASHRAE Standards, please do voice your opinions.

To answer the original question by Chris-

One may purchase the last years weather data for a small fee at NOAA website which provides the mean monthly outdoor temperature or may use a TMY3 weather file to extract the hourly data for the month in question. Calculate the daily max and daily min and then calculate the average of daily max and average of daily min and then calculate the average of these two averages. One may use only the summer months and/or shoulder seasons (depending on the climate) for estimating the adaptive thermal comfort envelope, because that's when one would use natural ventilation for cooling. Once you have the adaptive comfort envelope for the month in question, use the internal operative temperatures during the occupied hours to determine if 80% of those hours are within the upper and the lower limit of allowable indoor operative temperature.

Furthermore, some researchers believe that adaptive thermal comfort is a much complex phenomenon and hence the adaptive comfort envelope be estimated using the running average of the past 30 days to calculate the thermal comfort envelope for a particular day in a month and some simplify that to only past one day. Please refer to the work by McCartney & Nicol (2002), Nicol & Raja (1997) and Humphreys and Nicol (1998).

Thanks.

Gaurav Mehta

sorry, I should have read my response before hitting the send button. The following text would require some change-

......use the internal operative temperatures during the occupied hours to determine if 80% of those hours are all the hours are within the upper and the lower limit of 80% of acceptability limit allowable indoor operative temperature.

Best regards,

Gaurav

Thank you Gaurav,

I used the average of all (300 or so) hours for each weather month of an

EPW file. Rationale being:

- The method is comparitive so shouldn't depend too heavily on absolute

weather data (like we do in the UK)

- Unofficial advice from a technical adviser at ASHRAE confirmed it

- ASHRAE person pointed me to same results on Weather.com with close

agreement

- I calculated using the "means from means" method and it was within 0.5

of a degree

From discussion, a drawback appears to be that certain weather sets and

regions may unfairly load the result in either direction.

Joe Huang suggested that the original "means from means" method was a

legacy from the days when weather stations reported daily min and max

only. Simulation methods have moved on since. When this was put to the

ASHRAE adviser he recommended that this was put formally to ASHRAE as a

change request.

I'm intrigued as to where the naming of 80% and 90% limits come from.

Like you say Gaurav, neither of them refer to 80% or 90% of hours. The

nomenclature appears to be entirely arbitrary. In light of your

correction to your first email, it also seems evident that they can be a

little misleading.

Best regards

Chris Yates

Chris, perhaps I can shed some light on this apparition for you. I was a

member of the ASHRAE 55 committee when the adaptive comfort method was

introduced. From data sets with unknown origins a group came up with

statistical evaluation of the results, the mean outdoor temperature. I

see some of the top weather statisticians have already replied and

wonder above wonder when you take the mean temperature, it is somewhat

lower than the design temperature most of us old fogey's are used to

working with. So mysteriously a 32C design temp gets slimmed down to a

26C mean temp and it fits the graph.

When we have to 'cook' the books in this manner we simply take the

monthly outdoor dry bulb temperature and divide this by the number of

hours. Now the next question that isn't explained in 55 is whether or

not the weather data should be taken for only the occupied period or the

whole month ? The plot thickens. A simple explanation on how to use

this methodology is whether or not the user is liable for the

calculations and designs. Imagine a litigation case, when asking the

engineer 'How did you calculates this?" well I took the mean of the

outdoor temperature and I...................... Case closed.

Peter Simmonds Ph.D.