Ambient Temperature

[czars]

Is there an accepted source for determining ambient temperatures for use with the NEC Article 310.15?

 

[charlie b]

There is not. Furthermore, there is not an accepted interpretation regarding the meaning of "ambient." By that I mean there is some debate regarding whether the term refers to the highest mid-summer's day reading on record in the area, or the average daily highs for that area, or the year-round average, or some other value. I support the "year-round average" interpretation. There is some science behind that viewpoint, but I know of no formal, legal, or official statement (at least not in my area) that either endorses or opposes it.

 

[Dennis Alwon]

Here is a source that may be helpful, fwiw. Copper.org

 

[rbalex]

Depending on the Edition, the FPN in 310.15 B(*)(C) Conduits Exposed to Sunlight on Rooftop suggests ASHRAE Handbook — Fundamentals; however, all the problems charlie b mentioned exist because it is still ill-defined.

I am currently developing a Proposal to use a specific ASHRAE or possibly National Weather Service “design” percentage. These relate to the percentage of 8660hrs/year the ambient exceeds a specific temperature.

Edit add: "*" above is "2" in the 2008 NEC and "3" in the 2011 NEC.

 

[czars]

Thanks for responses guys. Sounds like an inclusion or a specific reference in the NEC is needed.

 

[weressl]

Depending on the Edition, the FPN in 310.15 B(*)(C) Conduits Exposed to Sunlight on Rooftop suggests ASHRAE Handbook ? Fundamentals; however, all the problems charlie b mentioned exist because it is still ill-defined.

I am currently developing a Proposal to use a specific ASHRAE or possibly National Weather Service ?design? percentage. These relate to the percentage of 8660hrs/year the ambient exceeds a specific temperature.

Edit add: "*" above is "2" in the 2008 NEC and "3" in the 2011 NEC.

Perhaps a defined time period duration of averaged maximum, not unlike how transformer or motor insulation life effect calculated. Ex. highest 4 hour average should not exceed 40C. That works for ambient where only the sun and atmospheric conditions dominate, if other heating effects are calculated you also integrated the rate of rise as relates to maximum temperature reached.

 

[Besoeker]

There is not. Furthermore, there is not an accepted interpretation regarding the meaning of "ambient". I support the "year-round average" interpretation.

I don't.
You can't rate equipment based on an average annual 10C (50F) when you might have a week of 25C (77F).

 

[weressl]

I don't.
You can't rate equipment based on an average annual 10C (50F) when you might have a week of 25C (77F).

I doubt there is a week of 25C (77F). There is day and night, so there is SOME variation even within the day, not to mention week.

 

[charlie b]

You can't rate equipment based on an average annual 10C (50F) when you might have a week of 25C (77F).

You can if you are talking about a conductor's insulation system, and if you apply the Arrhenius Equation using the times, temperatures, and insulation system physical properties. That is the technical basis for my opinion. I note, however, that this has not been incorporated into the NEC, and so therefore it has only the status of an opinion.

 

[Besoeker]

I doubt there is a week of 25C (77F). There is day and night, so there is SOME variation even within the day, not to mention week.

Well yes, variation between day and night. I've been places where it did get down to 25C at night.
My point is simpler. The rating of say, and electric motor, is ultimately limited by how hot it gets. Overload it for a period, it will get too hot and insulation failure will be the most likely cause of its demise. How hot depends on ambient temperature all other things being equal. Even if you have a a wide diurnal temperature variation, you still need to take account of the hotter periods.

Most, if not all, electrical equipment is temperature rated.
In UK, we have BS7671 (affectionately known as the electricians bible or just "the regs"). This has tables for conductor current carrying specified at a 30C ambient. There are also tables of correction factors for different ambient temperatures.

 

[Besoeker]

You can if you are talking about a conductor's insulation system, and if you apply the Arrhenius Equation using the times, temperatures, and insulation system physical properties.

If you have to take account of the rate of a chemical reaction.......hmmm

 

[weressl]

Well yes, variation between day and night. I've been places where it did get down to 25C at night.
My point is simpler. The rating of say, and electric motor, is ultimately limited by how hot it gets. Overload it for a period, it will get too hot and insulation failure will be the most likely cause of its demise. How hot depends on ambient temperature all other things being equal. Even if you have a a wide diurnal temperature variation, you still need to take account of the hotter periods.

Most, if not all, electrical equipment is temperature rated.
In UK, we have BS7671 (affectionately known as the electricians bible or just "the regs"). This has tables for conductor current carrying specified at a 30C ambient. There are also tables of correction factors for different ambient temperatures.

I think we are in general agreement; the question remains to define what the temperature means. The thermal aging is dependent not only on temperature, but also on the duration of that temperature extreme. Not to mention that the heat exchange is also affected by the humidity. The effectiveness of dissipation also depends on the mass and the surface area available for dissipation. What I am looking for is a datum point with two parameters; one is temperature expressed in degrees, the other is the time over which this temperature is integrated. Certainly humidity can also be part of the calculation(s) or an adjustment factor.

 

[charlie b]

If you have to take account of the rate of a chemical reaction.......hmmm

The degradation of a conductor's insulation system as a result of elevated temperature is a chemical reaction. My former professor and thesis advisor has done a lot of research on the subject. I even co-authored a paper with him:

A. K. Behera, C. E. Beck, and A. Alsammarae, “Cable Aging Phenomena under Accelerated Aging Conditions,” IEEE Transactions on Nuclear Science, vol. 43, no. 3, pp. 1889 - 1893: 1996.

 

[HoosierSparky]

In the Arizona desert communities, Phoenix, Scottsdale, Tucson, etc., ambient temperature is a big concern, escpecially now when so many PV installations are going in. During our peak temperature periods we have a high temp of 120F (measured in the shade). When you interpolate this when measuring the temps on rooftops and specifcally wiring in metal conduit, you can have some extremely hostile situations. We tend to err on the side of caution and use 61-73C correction factor from Table 310.16. When reading 310.10 it could be argued to go to the next higher.

It is not only the heat possibly generated by the current, but also the effect of the heat baking the insulation and possibly tempering the metal of the conductor changing its properties. When you start to consider all the factors, it quickly goes beyond my skill set to determine.

 

[Besoeker]

What I am looking for is a datum point with two parameters; one is temperature expressed in degrees, the other is the time over which this temperature is integrated.

The temperature expressed in degrees is fairly straightforward.
Duration is another matter.
A 5 MW motor might not be particularly adversely if the ambient temperature went 20C above it's rating for say, 10 minutes. I'm not sure a transistor in a TO5 case would be quite so tolerant.