Go Back

UL & NEC CONDUCTOR AND CONNECTOR TEMPERATURE RISE
THE CAUSES OF HEATING AND COOLING AND AMBIENT AMPACITY COMPENSATION.

NEC TABLE 310-16

A good foundation of the basic principles of heating and cooling in wires and connectors is helpful in seeing how the heat inputs are generated and how the heat is dissipated to produce the final wire or device temperature that meets the code, standards and reliability requirements of the particular equipment.    

 

Q.  What does the number 7 or 9 mean on UL or Canada marked wire connectors and wire?
 

A. This is shorthand for the temperature ratings that UL and the NEC use to make sure that connectors, electrical devices, wire and plastic insulation do not get hotter than intended in use. It is all about making sure that no part gets hotter than it should in the interests of longevity, safety and fire prevention.

 

7 is shorthand for 75C, 9 is shorthand for 90C. The operating temperature of a connector or wire is not to exceed that temperature limit.

A. The common typical mark is “CU7AL” or “AL7CU” meaning that copper or aluminum wire may be used at capacity based on a likely highest working temperature of 75C. CU9AL means 90C rated.

Q. Why does is matter if copper or aluminum which both have high melting temperatures get hot?

A. The temperature ratings are most important for protecting non-metallic materials like the plastic wire and connector block insulation. Plastics have much lower resistance to heat than the metals mostly used for wire, copper and aluminum. Worse, plastic can change its properties over time so what might seem fine in a short term test, may fail months or years later when the plastic has deteriorated from excessive exposure to heat and moisture.

UL and IEC standards are extremely rigorous about testing all insulating plastics for long term maintenance of their essential electrical and strength properties.

The reliability issue is not confined to plastics alone though. The contact reliability of an all metal wire connector is also related to temperature extremes that it sees in use, since expansion and contraction of dissimilar metals can eventually lead to a reduction of the gas tight (air tight) connection joint, leading to oxidation and a high resistance thermal break down of the connection.   

So stress testing for connectors is also based on how high the end use temperature is going to be.

 

Generally 90C rated connectors are more “massy” then 75C connectors since they need to dissipate more heat than one running in a 75C application and meet the stress test over current requirements of the “master” connector standard, UL486A-B.

Running a 90C connection requires a beefed up connector just as running 90C wire need improved insulation plastic.

 

Bulkier connectors have more surface area with which to transfer heat to ambient air.

 

As wire and connection temperatures increase, like say in heating element connections, materials change to high temperature metals like nickel plated steel, and high temperature insulators like fired ceramics, and ceramic or glass fibers, or in old installations, asbestos.   

Q. What limits the maximum allowed current in a conducting path?

A. The lowest temperature rated ‘link’ in the current ‘chain’ or the lowest “amp rated”  component’s listed current limit in the ‘chain’, the sets the allowed maximum current for a given gauge of wire.

Also the expected ambient temperature is taken into account since warmer ambient air reduces the heat dissipation rate. To compensate for higher ambient temperatures, the allowed current is reduced so that the current heating effect is reduced accordingly. Current is the leading factor in temperature rise because Watts = Current SQUARED X Resistance in the wire and connections.  That is why NEC de-rating tables are used to lower currents for conductors used in ambient temperatures above 30C. 

 

These so called “capacity” (amp capacity) tables are based on one base value. If the surrounding temperature is higher, then the same wire gauge and conductor material with the same current will reach an overall higher temperature so the allowed current is scaled back by a scaling factor from the table. Tables with these scaled back capacity numbers are available in the NEC code book.

Q. What are the factors behind wire getting to this temperature limit?
     
A. The temperature of a given gauge of wire depends mainly upon the resistance of the wire and the current flowing through it.

The bigger the gauge of the wire, the more current can flow though it and it remain at a satisfactory temperature.

Wire gauge designations, AWG or kcmil / MCM or mm2 are all tied to the cross sectional area of the wire, not the diameter of the wire. Because resistance decreases in direct proportion to the cross sectional area of the wire (think about two parallel wires like two parallel resistors, so if each the two wires has half the cross section of a single wire, both cases single and double wires will have the same overall resistance.
 
Below, the reason for the two wires case running cooler than the one wire will be discussed.  

Q. Apart from ambient temperature what other factors help cool the wire in “still” air?

A.  Stranding class and insulation type affects the total outside diameter of the wire. Finer stranded wires have more air between the twisted or woven strands so take up a larger space than solid wire (no air) or even coarse, rigid stranded wire (medium air).

The largest commonly used fine stranded flexible conductor is class K, often used for welding cables and is noticeably larger than standard code (rigid stranded) wire class B or C stranding.

A benefit of greater diameter is that greater air cooling will occur. The least cooled wire is a non-insulated solid wire since it has the least surface area in contact with ambient cooling air.

It is rather counter intuitive that an “insulated” wire would cool better than the non-insulated wire. The reason is that the weakest link in cooling the wire is the interface with the ambient air which benefits more from increased surface area of thicker, larger diameter insulation than the thermal resistance of the insulation. 

The large class K wire with the thickest insulation on it will tend to have the greatest cooling benefit in contact with ambient cooling air.    

 
While wires benefit from having a larger diameter for improved cooling contact area with ambient air, the math will show that as the wires get larger the contact area ratio with cross sectional area falls off because wire cross section increases with radius SQUARED, whereas the contact area with air increases with radius only LINEARLY.

Consequently, the current capacity of wire, per area of conductor, falls off as wires get larger, which has led to a practice of paralleling two, or three smaller wires rather than one big one.

The air contact surface area is thereby increased. An adverse offset of having three wires running together is some mutual heating going on, but overall, there is enough to be gained by this paralleling practice that it is widely used. UL only permits it for wires of 1/0 AWG and larger since the benefits are less for smaller wires and there is an increased risk of avalanche failure if one wire connection fails, the others have to take up the slack. The other benefit of being able to bend two or three smaller cables easier than one big one is also a big advantage.  

Q. How do you interpret the NEC capacity Tables?

A. The three columns in the 30C ambient capacity chart, 60C, 75C  and 90C are not  the temperature  the wire will become  at the posted current values but  it is predicted that it will not get hotter than 60C, 75C or 90C. These charts were based on empirical test data in still air and have been the foundation of heat rise guidelines since definitive and comprehensive testing was done starting in 1938. 

 

The name of the insulation (say for example TW), and the fact that its name appears at the top of the 60C column, says that  the wire can handle an overall temperature of 60C, without suffering damage to its insulation. THHN insulation on the appropriate gauge wire for the listed current will be able to withstand 90C.

Q.  So if 90C higher temperature insulation is readily available why don’t we use the thinnest wire we can to carry the current to save copper costs?

A. UL508A is the ubiquitous standard for panel builder shops and sets the currents based on 75C as the maximum permitted, as a conservative design parameter for panels having a lot of heat generation within a tight space. Consequently, many manufacturers have not yet listed many of their connectors, circuit breakers and switches for more than 75C. That is why generally capacity higher than those listed in the 75C column are not used, even if the wire now commonly has 90C insulation as standard.

One of the complications of attaching wires to fuse blocks and circuit breakers is that the heat in the wire will tend to affect the temperature of the sensing elements in the circuit protection device and so it is important for consistent performance to limit the variation of heat flow that could otherwise occur.  

Increasingly electrical assemblies outside of the UL508A domain are using 90C or even 105C and taking advantage of higher capacity provided by the higher temperature insulation. 

Q.  So, there is no harm in using a 90C device or wire in a 75C regulated application provided that the currents used are from the 75C capacity column?

A. Correct. This would be perfectly acceptable in the same way that using heavier gauge wire then needed is acceptable, all other thing being to code.

Q.  When you can use 90C capacity, does it always make sense to do so?

A. Generally yes but there is a tradeoff. Lower copper costs of smaller gauge wire for a given current is offset by higher I2R losses in watts of wasted power getting that current where it is needed to do real work.

 

Q. When you have one part rated at 75C and the other at 90C and they’re mounted together, does that affect the entire assembly temperature rating?

 

A. Yes. Each part has to carry the labeling in the file under its catalog number and match the markings on the appropriate connector, but when multiple UL parts and UL wire are assembled into a current path, the lowest temperature rated conducting part is the one used to calculate the maximum current allowed under the temperature rise formula. It’s the good old “weakest link in the chain” methodology with a twist since there are multiple smaller “chains” on the many 4-14 AWG wire holes.

 

All xS4-3 parts should be described as CU7AL current ratings with or without the S350 lug accessory, but since the 75C current of many #4 wires load wires can far exceed the current rating of the S350, the maximum current allowed on a single S350 should be as though it were rated at 75C.

 

When bolted to a properly sized conducting bus the entire current rating of 24 x #4 or 36 x #4 wires is available but if used in a power distribution (or power collection) mode (no bus) with the S350 (350 kcmil) as line lug, then the 350 kcmil wire is current limiting at 75C ampacity.

 

Listed Wire Connectors used as grounding blocks (insulated floating grounds or neutrals or bolted to case grounds) where permitted, such as UL508A industrial panels, have their own rules in the standard on wire sizing for ground currents versus when using in power circuits.

Q.  What if you have hard-to-get solar power current that might have to run a long distance back to the place where power is accumulated and inverted to AC for use?

A. Using a heavier gauge wire is often appropriate to reduce I2R losses by decreasing R, resistance. Since the cost of copper is high going to an even more oversized aluminum wire can be very economical since the cost of the bigger wire is less than smaller copper wire and the I2R watts losses are lowered. This strategy is particularly effective for long distance runs of wire.

This accounts for the popularity of so called “Dual Rated” mechanical wire connectors that can terminate either copper or aluminum wire and also “range take” wire gauges of different size and capacity to control the watts loss rates per foot.

 

Q.   How do the de-rating factors for current work when the ambient is higher than 30C?

A.  For example the capacity of a 250 MCM wire made from aluminum with say XHHW-2 insulation, is 230 amps.

At ambient temperature of 112F, the current is to be reduced to prevent it from rising above the 90C that the insulation is rated for. The de-rating factor is 0.87

0.87 X 230A is 200 amps

Q. what other factors are involved in capacity of wires?

 

A.

1. The number of wires in a wire way.
The number of wires that run together since the presence of other hot wires in the same wire way will lower the ability of each wire to cool in ambient air. The NEC has rules on multiple wires (typically more than 3) that complicate the final capacity of each wire so there is much to learn from the NEC Code beyond the basics in this FAQ 

2. Forced Air cooling
In moving air applications (fan or chimney convection forced air cooling) different rules apply. This is more common in electronics where the controls to roll back or cut the current can be built in to prevent overheating. 

 

3. Frequency
In AC applications 60 or 50 Hertz are close to DC for wire heating purposes if there is no capacitance or inductance reactance in the wire chain.   
60 Hz is the default safe test frequency since any capacitance or inductance reactance will be higher at 60 Hz than 50 Hz.

A new heating factor is now common with PWM (Pulse Width Modulated) currents used in switched DC variable speed or power controllers.

There are PWM power supplies that can use 100K Hz to chop up the DC current and so there are significant reductions in capacity from heat rise from L and C reactance and skin effect changes in the conductivity of wire. Even the choice of wire way can result in unexpected inductive reactance since a steel tray can be the magnetic “core” to create high levels of inductance.  

4. Duty Cycle
On the other side slow frequency DC chopping like a 50/50 square wave has a lower overall heating effect than AC RMS (sinusoidal). If the full current is never seen for continuously then up rating can occur since the effective “RMS” heating effect current level is lower.

 

5. High Frequency Harmonics
Steep wave forms like the near vertical rise and fall of chopped “square” waves from switching power supplies have higher frequency harmonics than their fundamental chopping frequency. The steep part of the rise and fall of a square wave mimics a sin wave of an extremely high frequency, albeit for a very short time.

 

It can be expected that this high frequency “noise” will add a little to the heat generated in a conductor since it will encounter the most impedance in the presence of even small amounts of inductance and capacitance in the current path.