20 amp circuit in 160?F attic

[paulshanks]

After reading through home inspection news articles, and hearing from inspectors through forums and directly through email, it is clear that attics in homes (at least in some areas) frequently exceed 140?F. These are people who actually have to go into multiple attics every day throughout the summer, and have equipment to measure and document temperature accurately. In fact certified inspectors have provisions which say they do not have to enter an attic when it is over 140?F. Certainly there is a difference between surface and ambient temperature, but those that are using IR say that there is only a +/-3?F (6?F total) difference between roof and insulation.


A wide range of "highest" temperatures have been reported. I will throw out the absolute highest temperature reported, which is 189?F - assuming it is erroneous. The next highest temperature of 180?F, I will assume is the roof, and if I double the normal 6?F differential between roof and insulation, I still come out to 168?F down around the electrical cables. Most inspectors report highest temperature of around 160?F, and at least one reported more than 140?F in 78?F outdoor weather (this is a differential of more than 60?F). Lastly, I have not yet found one home inspector who says attics "never" exceed 140?F in the Dallas area.


Obviously these temperatures are inefficient for HVAC and can cause damage to the roof, but most importantly, they create a fire hazard for 20 amp circuits on 12 gauge cable. Sure, it would be best to improve attic ventilation, but without knowing the condition of the attic, if it is possible for it to exceed 140, why install cable that would be dangerous in that situation? Stranger still, why doesn't the NEC have any provisions for it?


When I have pointed this out to electricians, most recommend using higher temperature insulation. I guess I don't expect them to have the NEC tables memorized, but these values are calculated using the highest temperature rated insulation, which is 90?C (194?F), and therefore, there are no other insulation options. I have had some non-electricians recommend 10 gauge cable. 10 gauge allows for close to 30 degrees higher ambient temperature than 12 gauge (at 20 amps, no conduit, 90?C rated insulation, etc.), making it suitable for most of the temperatures reported by inspectors. The problem is, 10 gauge isn't well suited for standard sized outlet boxes, receptacles or switches. So, using 10 gauge cable on these devices would not meet code.

In summary:
Attics do exceed 140?F
Better insulation is not an option
10 gauge cable may not be an option (for general lighting and receptacles)
What is the answer?

 

[George Stolz]

Install the NM under the insulation. Standard practice.

 

[Ragin Cajun]

Using my own home as an example built in 1967, I haven't seen any evidence of insulation issues in the attic. Ditto for other attics I have had to be in.

The only place in a residence I have seen briddle insulation is in a light fixture outlet where a much larger lamp was used or at a bad connection or defective receptacle.

I wonder if the insulation really is better than the code lets on?

Also, just how many home branch circuits really see heavy loading over an extended period of time? Diversity is our friend. Compare the code "demand" numbers to those used by the utilities, BIG difference!

RC

 

[infinity]

Install the NM under the insulation. Standard practice.

Agreed, then it will operate in an ambient near that of the living space below.

 

[iwire]

When I have pointed this out to electricians, most recommend using higher temperature insulation. I guess I don't expect them to have the NEC tables memorized, but these values are calculated using the highest temperature rated insulation, which is 90?C (194?F), and therefore, there are no other insulation options.

I am confused, NM is rated to be installed in 90C areas which is less than the highest temp you are reporting.

I have had some non-electricians recommend 10 gauge cable. 10 gauge allows for close to 30 degrees higher ambient temperature than 12 gauge (at 20 amps, no conduit, 90?C rated insulation, etc.), making it suitable for most of the temperatures reported by inspectors. The problem is, 10 gauge isn't well suited for standard sized outlet boxes, receptacles or switches. So, using 10 gauge cable on these devices would not meet code.

It 'meets code' we use 10 AWG on devices often.

Yes, it may mean you have to use larger boxes.

In summary:
Attics do exceed 140?F
Better insulation is not an option
10 gauge cable may not be an option (for general lighting and receptacles)
What is the answer?

Run larger wire or don't run the wire in the hot space.

 

[dhalleron]

I used to work in a lot of houses that had no insulation in the attic. Many had a few inches of soot that leaked from the chimney. My mom used to bleach my white tee shirts until they fell apart over 30 years ago.

I often ran into the old BX or AC cable that had been in there for many years. The insulation on the wire in the fixture box was brittle from years of high wattage bulbs being used. We used to break back the armor and find good wax covered cloth insulation inside the attic. Then we would set a box and run new wire back into the light fixture.

None of those circuits carried much of a load, so that probably helped. Plus in Kentucky it doesn't often get above 100 outside. I have been in an attic when it was 100 outside and I didn't stay long.

 

[Dennis Alwon]

We live in NC and the attics here get hot-- not sure I have ever seen one at 160?(F) but they can get 130?.

That being said I will reiterate what has already been said. In a home the load on conductors are usually not for extended periods of time so the heat does not appear to cause issues. The air temp is never 160? all year long so I wonder if it is something to worry about. The NEC mentions ambient temp but the meaning of ambient makes this rule very difficult to use. There are websites that give temperatures ranges for different areas of the country but I would not call that ambient temps. Sorry I wasn't much help was I?

 

[GoldDigger]

I am confused, NM is rated to be installed in 90C areas which is less [-->more, ed.] than the highest temp you are reporting.

But even if it is allowed to be used in 90C ambient, it will then have zero tolerance for I²R heating which would take the insulation temperature higher.

 

[paulshanks]

Thanks for the responses

Thanks for the responses

Thanks for the responses!

I realize that this situation is unlikely. 160?F would be, at most, only part of the day, and only part of the year. Also, it is not likely that the circuit will have a full load on it for an extended period of time during those temperatures.

Ragin Cajun, you are probably right that the insulation is better that the code lets on. My guess is they have some tolerance. The problem here is that we don't know how much tolerance, and it only exists when ambient temps are at or below 140. If the ambient temperature is above 140?F, some or all of the tolerance may be negated.

Iwire, as GoldDigger points out, at 140?F of ambeint temperature, the wire starts out at 140?F with zero current. If you then add 20 amps for a period of time, this raises the wire's temperature to 194?F (90?C) which is the temperature limit of the insulation.

10 AWG may be the remedy if I use a junction box to connect the 10 AWG cable from the ceiling to the 12 AWG in the wall; or (like Iwire said) install larger outlet boxes. Then I guess I could use 12 AWG pigtails to ensure a secure connection at the receptacle. 10 AWG also would run 70 feet before resulting in a 3% voltage drop on the branch (12 AWG would only run 45 feet). However, since this is rarely done, I feel like I am missing something. Some say it is a waste of copper, but is seems like a logical solution to meet the NEC ambient temperature and voltage loss recommendations.

Any other thoughts? Thanks again!

 

[George Stolz]

10 AWG may be the remedy if I use a junction box to connect the 10 AWG cable from the ceiling to the 12 AWG in the wall; or (like Iwire said) install larger outlet boxes. Then I guess I could use 12 AWG pigtails to ensure a secure connection at the receptacle. 10 AWG also would run 70 feet before resulting in a 3% voltage drop on the branch (12 AWG would only run 45 feet). However, since this is rarely done, I feel like I am missing something. Some say it is a waste of copper, but is seems like a logical solution to meet the NEC ambient temperature and voltage loss recommendations.

Any other thoughts? Thanks again!

It is becoming increasingly clear that this is not a case of checking up on the practice of a hired electrician, but in actuality DIY. I am closing the thread.