"As I mused, the fire burned"

Reflection on life as a person of faith.

Derating of Electrical Cables in Thermal Insulation

with one comment

I submitted this to the Canadian Standards Association (CSA) as a proposed amendment to the Canadian Electrical Code. The subcommittee concerned with the code rejected the amendment unanimously.

I published (along with two collegues) an article in a trade journal on the subject. This was my attempt to get word out to electrical professionals that there were greater issues involved when installing electrical cables into insulation. In most Commonwealth countries, their electrical codes have been modified to include derating factors since about 2008. In the UK, a cable surrounded with thermal insulation must be derating to 50% of its free-air current carrying capacity. I’m presently dialoguing with an Australian architect who is examining the derating of electrical cables installed in masonry walls filled with thermal insulation. This is a present issue, and I was a little shocked that the CSA dispensed with the proposal.

The reasons? There was no evidence of a problem in the field (although my fire investigation was exactly that). The present code was clear (our rule 2.122 just says be cautious, while the UK code has specific derating factors). There was no need perceived (although the UK, Aussie and Kiwi codes have all been changed…along with the USA NEC in more recent days).

The reason we don’t see this problem frequently is because most residential installations are done with a great enough safety factor (over design), that the circuits are rarely operated near capacity. So a 15 amp lighting/outlet circuit will rarely be operated at even 50% of that load. I have some future concerns – particularly with the popularity of spray foam insulation. I have seen installations where the foam is sprayed over the installed electrics, encompassing the electrical cables fully. The thickness of the spray is dependant on the operator. That high safety factor will protect most of those installations, but might not cover an electrical installation that is running near capacity.

As an example, consider a standard 15 amp circuit installed in a garage or shed by a do-it-yourselfer to service wall outlets and lighting (I’m not considering the code provisions for installation here, just to illustrate the loading).  The owner then uses the circuit to run lighting, as well as a forced-air space heater operating constantly to warm the garage/shed.  During the build he insulated by spraying foam insulation into all the voids between joists, and did this after the wiring was installed, so the cables are completely surrounded in thermal insulation.  Now, the reason we don’t see these failures happening on a daily basis is because: the ampacity tables in the code already have a hefty safety factor built into the values; almost no home-owner runs a circuit near capacity for any length of time (how long does your kettle, or hair dryer operate over a 24 hour period?).  In fact, there is effectively no heavy-draw appliance that would be classed as continuous operation in a residence (the code says operating at least one hour out of every two hours).  In my investigation, the cable that failed was operated near capacity (200 amps versus 215 amps) and was a continuous load cable.

So, the approximate math:

A 1500 watt space heater, on a 120 volt circuit, draws (Ohm’s law: V=IR; P=VI=I^2R) a current of 12.5 amps.  The home owner also runs three 100 watt light bulbs on the circuit, each drawing 0.83 amps for a total of 2.5 amps.  The circuit is now operating at capacity of 15 amps.  If the heater is left operating with the lights off, the circuit would still be over capacity according to the British electrical code, which would limit the circuit to 7.5 amps.

If that cable is completely surrounded with thermal insulation for a good distance of its run, it will eventually fail due to over-capacity use.  This is one of the reasons grow-ops burn, because they are one residential application where the circuits are loaded to and beyond capacity on a continuous basis (aside from all the unlawful stuff).  This is also the reason to be very cautious about operating space heaters on a continuous basis.

I found this perplexing comment at a USA spray foam website:

Can Spray Foam Insulation be sprayed over electrical wiring?
Spray Foam insulation does not pose any issues with electrical wiring as long as the electrical wiring is installed per National Electric Code. Any overheating issues with Spray Foam insulation in regards to wiring would be directly related to improper installation of the circuit or the size of the wiring. Polyurethane foam is chemically compatible with electrical wiring plastic.

So…if you install the wire correctly by the NEC (that is, you properly derate the wire for the impact of the thermal insulation)…the wire won’t overheat…but if you don’t install it correctly, and it overheats, its not the fault of the foam insulation, it’s the bad installation of the wire.  I obviously spend too much time thinking like a lawyer: but for the thermal insulation, the cable would not have been overloaded.  So what causes the overload?

Anyway, here’s my submission – just so I can say I noted it first when the Canadian code is finally amended.


Introduction:  I am writing with a suggestion for an addition to the Canadian Electrical Code (CEC) Tables 1, 2, 3 & 4 and the related rules.  These tables concern the allowable ampacities for copper and aluminum conductors in various configurations.  This suggestion also includes rule 2-122, Use of thermal insulation.

The motivation for writing this suggested amendment arose out of a forensic fire investigation I conducted as a consulting electrical engineer.

Incident Description:  An electrician installed a 250 kcmil armoured aluminum 3-conductor plus neutral cable (ACWU90) in a food processing plant to power a piece of process equipment.  The cable was routed through an attic space and was initially laid on top of the loose thermal insulation layer of about 16 inches depth for a total run of about 120 feet.

The cable was providing 3-phase 208 VAC power to a transformer through a shut-off switch which included 200 amp phase fusing.  The transformer in turn fed the process equipment.  The cable originated at the building supply panel where it was protected by a 225 amp moulded case circuit breaker.  After installation the electrician carried out a worst-case measurement of current indicating the maximum load was slightly less than 200 amps per phase.  The cable had been installed for about one year.

After operating the process machine the plant was left empty for the weekend.  About 12 hours after the process machine had been shut down the owner received a call from the building alarm service that there was a power failure in the building.  When the owner arrived he found the circuit breaker for the process equipment and the building main breaker tripped.  He reset the building main breaker but when he attempted to reset the process equipment breaker it immediately tripped off.  The owner left.  When he returned approximately 6.5 hours later to open for the day he found the building full of smoke.

When the fire department began suppression they found a smouldering fire involving the roof trusses buried in cellulose and rock wool insulation in two separate locations.  The point of origin had been nearly simultaneous at the two points which were both along this cable run.  The cable neatly and almost perfectly bisected each of the areas of fire damage.  The cable armour was burned away and melted for about one foot at the centre of each fire area.  A third location had become hot enough to melt the outer thermoplastic jacket.  There was evidence of electrical arcing coincident with the areas of armour damage.

Investigation:  The process equipment was a continuous load device as it was operating at least 1 hour out of every 2, under rule 8-104(3)(A).  For the cable installed, Table 4 provided an ampacity of 215 amps for 90° C cable.  Given the circuit fusing at 200 amps, this appeared to be a code-compliant installation.  I could find nothing involving the process equipment or the transformer that suggested a cause of the fire.

The two areas of worst fire damage showed clear evidence that portions of the armoured cable had been in contact with the ceiling joists.  There were burned-out notches in the joists that matched the cable armour dimensions.  This suggested the cable at both of the points of origin had been buried under about 16 inches of loose thermal insulation.  In the third area, where the thermoplastic insulation had melted off the armour, the cable had only been buried under 1 to 2 inches of the loose thermal insulation.  Other areas, where the supply cable remained resting on the thermal insulation, displayed no damage from heat.

The cause of the fire was identified as electrical overheating of the supply cable.

The only rule in the CEC concerning de-rating the ampacity of electrical cable which may be buried in thermal insulation is 2-122(1)(a):

 …special care shall be taken to ensure that conductor insulation temperatures are not exceeded due either to mutual heating of adjacent conductors or cables or to reduced heat dissipation through the thermal insulation…

A literature search provided British Standard 7671:2008.  An addition to this electrical code was issued in Regulation 523.7 which states that a de-rating factor of 0.5 should be applied to the current rating of a cable if it is completely surrounded by thermal insulation.  Applying the BS7671:2008 de-rating rule would have resulted in that cable installation being limited to 107.5 amps, just over half of the measured current flow in the cable.

The cause of the electrical overheating of the supply cable was found to be the result of its burial under the 16 inches of loose thermal insulation.

While the supply cable had initially been laid across the surface of the attic insulation, the residual twisting of the cable forced two portions into the attic insulation.  Over a period of time, that residual twist caused the cable to settle into the loose thermal insulation, permitting an overheating of the supply cable and eventual failure.

The installation electrician was sceptical when we discussed these findings.  After I presented him the analysis, he was shocked to find that what he considered to be a code-compliant installation, which had passed inspection by the municipality, had been the primary cause of the fire.  I pointed out Rule 2-122, which he commented offered him no guidance on how to accommodate thermal insulation.

There should be some indication in the CEC as to the hazard of placing electrical cables in a manner that may result in them being later buried in thermal insulation.  The British amendment arose at least partly out of some testing as to the impact of adding attic insulation to older structures.  I have enclosed the article that discusses that testing.

Recommended Amendment:  By way of an amendment to the 2009 code add a specific warning to each of Tables 1, 2, 3 & 4 to caution about the need to de-rate if a cable is surrounded by thermal insulation.  While Tables 1 & 3 specify that the ampacities refer to a conductor in free air, Tables 2 & 4 refer only to a raceway or cable and do not include the free air delimiter.  I would also suggest that the words ‘free air’ in the title are not sufficient to inform an electrical technician of the inherent limits in those tables.  A suggested text:

The ampacities listed in this table are only valid for a cable in free air.  If any portion of the cable is surrounded by thermal insulation, the ampacities need to be de-rated.  Failure to de-rate for thermal insulation can result in cable failure and fire.

For the next edition of the CEC, it is suggested that Rule 2-122 be amended to reflect specific guidance on the de-rating of cables surrounded by thermal insulation.  This will likely require the addition of ampacity tables for this condition.  The research conducted by the United Kingdom may be a useful starting point for development of this revised rule.

Update February 2016.  Picked this up for an industry magazine.  First comment is from NEMA.

The second is from the chief electrical inspector in Washington.

They both suggest there is no problem.  Most interesting to me is the U od Toronto study from 1985.  It seems that the problem would be highly dependent on the thickness of the spray foam, and the number of conductors, and the total loading on the circuit.  Without knowing how the study was done, it’s hard to say the problem does not exist.

I am tempted to run this test myself.  It would not be too costly to set up a number of conductors, with a layer of spray foam, then load and monitor temperature.  There should be a nice graphical relationship between insulation, load and temperature.




Written by sameo416

January 18, 2013 at 11:05 am

Posted in Uncategorized

One Response

Subscribe to comments with RSS.

  1. […] A Wonderfull Mess I believe you are right to be concerned about the electrical wires as well. So is this guy. Apparently it is not a code issue (yet). If the circuit is overdesigned it will never see its […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Urbane Adventurer: Amiskwacî

thoughts of an urban Métis scholar (and sometimes a Mouthy Michif, PhD)

Joshua 1:9

Reflection on life as a person of faith.

Engineering Ethics Blog

Reflection on life as a person of faith.


Today, the Future and the Past all kinda rolled up in one.


For Those Courageous in Standing for Truth


Law, language, life: A Plains Cree speaking Métis woman in Montreal

Malcolm Guite

Blog for poet and singer-songwriter Malcolm Guite

"As I mused, the fire burned"

Reflection on life as a person of faith.

%d bloggers like this: