Lowest One Day Mean Temp - Canada

[Samm17]

In the National Building Code of Canada the lowest January temperatures are given for locations across Canada. These low temperatures are reported as having a 1% chance of being exceeded. Does anyone know how this would compare to a lowest one day mean temperature?

 

[JStephen]

The chart in API-650 extend a fair ways up into Canada, are you off the chart?

I see online information about average lows, highs, etc., in different cities, I would think you could compare those to the LODMT and make a reasonable approximation for areas off the chart.

 

[Duwe6]

I would have no problem using that data, Samm17. The engineering constraint is the Lowest One-Day Mean temperature. It is not a 100 or 200-year low that you are designing to.

 

[JStephen]

Note that the LODMT should be higher- the lows would be sometime before dawn, the LODMT would be averaged over a day (or possibly, the average of high and low temp for the day)- and using lows only might be unduly conservative.

 

[Duwe6]

I'd still be comfortable using the temp's from the Canadian Building Code. First, they are government-supplied and unassailable in court. Second, when you get 'way up north' the sun doesn't really rise during winter so the overnight low and the "noon-time" high temp are pretty close together. It's a whole 'nother climate up there.

 

[Samm17]

Thanks for the input.

I have seen the chart in API 650 that shows the temperature ranges in Canada but it doesn't agree very well with the published data in the National Building Code. (The NBC tends to have a warmer low temps than API 650. Example, Edmonton Alberta NBC -29F VS API 650 -43F). I have always designed to the NBC as it is the national standard and Fig 4.2 hasn't been updated since 1952. This leaves me a bit open to criticism as I am not designing to the worst case standard.

 

[geesamand]

Throughout the US, I see a significant amount of PVQ plate material available pre-tested for API 650 requirements of -40 MDMT. Depending on the temperature and reference you choose that may be conservative. In any case that drives my material selection when we build tank flanges - it's not worth it to me to try and source anything else because this is what's available.

To the root of the question, the two sets of data are fundamentally different. Fundamental to the 1% statistic is the duration of the probability estimate. If it's 1% chance daily, then that means 3 days per year below the number is normal. If it's 1% chance of occurring over a 50 year lifetime for the building, then it's much less frequent.

David