How to experimentally calculate power cable ampacity

[EE_HV]

Hello,

My online research has revealed international standards to calculate steady state ampacity of power cables (e.g. IEC 60287), as well as some software programs (Lineamps) and FEA calculations to predict thermal resistance and ampacity of power cables.

My question is how to experimentally determine the ampacity of a power cable, so that I can be confident that I am using the a cable gauge that is not going to fail/overheat or have a reduced lifetime.

I will be using high voltage (up to 400VDC, 400V AC) power lines to conduct between 200 and 400 amperes (DC and AC). The power cables are shielded (for EMC shielding).
The cables will be providing power to DC/AC inverters, that will in turn be attached to an electric motor.

The reason I want to determine this experimentally, is to account for the realistic environmental conditions, including heat loss due to conduction, without having to create a complex mathematical model.

My idea is to attach thermal probes to the cables (or use infrared meters) while running the inverter/motors assembly, while being subjected to different ambient temperatures and mechanical loads.
Temperature measurements will be taken until stable tempeature is reached, and then a different load or ambient temperature will be set.

So I will be able to create a table with the cable's insulation temperature, at different ambient temperatures and loads.

Once I have that information, how can I calculate the thermal resistance, and the ampacity of the cable?
What information I need from the cable supplier to determine this ampacity? (insulation temp rating, etc?)

Thank you very much in advance for your suggestions.

 

[MATLABdude]

You could probably save yourself a whole lot of work if you contacted a wiring supply company (or a manufacturer, e.g. Belden Cable or Alpha Wire) and told them what you needed. You could probably even get the tabulated data under various cooling conditions.

 

[EE_HV]

MatlabDude, thanks for the response.
I have contacted the manufacturer (Champlain cable), and the information will give me an idea of the expected ampacity, but the information they have will not take into account the actual operating conditions.
That is why I want to conduct a subsystem test, with the inverter and motor running at various loads and ambient temps. I may even want to simulate the airflow.

My main question is how I determine the ampacity with the experimental data.

 

[MATLABdude]

Okay, but is what you're doing so far out of normal usage conditions that you're unable to spec something based on the figures that they provide? Ampacity is usually determined under given cooling conditions (e.g. in insulation, water cooled, air cooled, still air, etc.) such that the temperature rise (above ambient) doesn't exceed a certain amount. Or, worse yet, melt.

To determine ampacity, you'd need to figure out what the maximum safe temperature increase can be (you could base this on the maximum temperature, and the maximum ambient temperature you expect to occur). Then you can measure the temperature (probably even just by hand) under various current conditions, assuming that you have basically still air (more air flow would lead to better cooling). And then, you derate it by a certain factor just in case you were too optimistic.

But if I understand correctly, you've already decided on a certain type of cable of a certain gauge, and presumably bought it. And now you just want to verify that it works? Otherwise, you'd need to determine how far away from test conditions you'll be operating at, and what impact that'll have, and go with higher / lower gauge as appropriate.

 

[EE_HV]

Ampacity from a manufacturer is normally determined at a rather low temperature, and no airflow.
So those numbers are not helpful, unless a derating table for temperature and airflow would be provided.

It is not that we are far out from normal usage, but rather that it is a dynamic situation so we need to test under a bunch of different conditions.
For example, under worst case load, we could actually have some decent airflow and the ambient temperatures could also be lower.

Also, the ampacity numbers from manufacturers do not normally take into account cable heat loses due to conduction, which at the lower temperatures could be significant (i.e. the motor and inverters could act as heatsinks). Or self heating of the inverter and motor could actually add heat to the cable.
So you can see there are so many factors involved, that the only way to find out the maximum cable temperatures is via experimentation.

Once I have the test data, does anyone here know how to translate that data to dermine if we are overheating the cable?
Is it as simple as looking at the cable's insulation ratings (e.g. 150C) and then make sure the max temp is not above that?
I am guessing not, since measuring outside temperature of say 150C means the copper touching the insulation is at higher temperature (e.g. 160C).

In other words: once I know the cables' worst-case outside temperature,what is the criteria used to determine if that temperature is exceeding the cables' ratings and not diminishing its life?

 

[dlgoff]

...what is the criteria used to determine if that temperature is exceeding the cables' ratings and not diminishing its life?

Maybe a fire? You shoud rely on the manufactures specs. and select one that is maybe 50% better to guarantee you don't burn the place down.

 

[EE_HV]

right, a fire would be a bad sign

seriously:
I want to avoid a rule-of-thumb where I just add a certain percentage to have a margin of safety.

How do the cable manufacturers determine the ampacity, other than the above mentioned IEC standards and related calculations?

They should be a bunch of experimental tests they do, from which then they can publish their guaranteed ampacity.

Anyone knows?

 

[Averagesupernova]

They pretty much determine it from known properties of the materials used to manufacture the wire/cable. They then factor in the amount of heat said cable is able to dissipate assuming it is inside conduit with other wires running at ampacity along side. These 'rules-of-thumb' are tried and true. One thing I don't think you have mentioned is how much voltage loss are you willing to tolerate. You may be willing to tolerate less loss than the acceptable wire size will give you. The NEC doesn't really care about loss, just safety. The tolerated amount of loss and the minimum wire size allowed for safety do not coincide especially where long distances are involved. If you run a considerable distance you will find the wire you need to prevent excessive voltage drop is much larger than the minimum required size for safety.

 

[EE_HV]

Thank you for the answer Averagesupernova.

I think the method you describe is along the lines of what IEC 60287 calculations do.
I would think that manufacturers also do experiments to confirm their calculations.
It is that part that I want to learn more about: how do they conduct their experiments and how do they then confirm that the wire is reliable to work at a certain amperage?

On rules of thumb: They are useful for quick assesments, but I have seen ampacity calculations of up to 2:1 for the same gauge, depending on who is publishing the information. That is why I am avoiding them for this case.

On voltage drop: it is a concern, but not a great one, since distances are relatively short (between 1 and 3 meters).

 

[Averagesupernova]

Most of the time differences in ampacity are because of WHERE the wire is installed. I have some older reference books dealing with codes and ampacities and in these books overhead wires run in free air individually, not several wires twisted together, can run double the amperage compared to wires in cable or conduit. This type of wire is not used anymore and has been replaced with what is typically called triplex.

 

[marcusl]

Good grief, don't try to do it experimentally. What criteria would you use for success? failure? How do you know how much safety margin to allow for? In short, why in the world do you think you can do a better job than hundreds of wire manufacturers, finite element models, and numerous standards and regulatory agencies?

Current capacity and temperature derating information is widely available. "Heat sinking" to the motor is a fallacy--your motors should run hotter than the wires, or else you've made a mistake. The IEC spec is probably right on. If you don't believe it, contact another wire vendor and ask to speak to an applications engineer. Ask for derating tables/curves for your application.

If you are worried about a lack of cooling, MIL-STD-975 gives safe current capacity and temperature derating for wires and wire bundles in the ultimate lack-of-cooling environment--a vacuum. See table 3.16 in
http://www.everyspec.com/MIL-STD/MIL-STD+(0900+-+1099)/"

 

[fizz_it]

This does not directly answer your question but it may help you determine your own comfort level.

From Circuit Cellar Issue 222 page 25 Author Steve Hendrix

The wire gauge is actually decibels (dB) relative to 100 μΩ per foot. That is, AWG#0 wire is 100 μΩ (0.0001 Ω) per foot. Each increase of 10 gauges increases the resistance by a factor of 10, so AWG#10 wire is 1 mΩ (0.001 Ω) per foot, AWG#20 is 10 mΩ (0.01 Ω) per foot, AWG#30 is 100 mΩ (0.1 Ω) per foot, and AWG#40 is 1 Ω per foot.

From this you can calculate the power dissipated in the cable (yes you could just measure it). Knowing you maximum ambient temperature and the thermal resistance of the wire, you could then determine how hot the wire is expected to get assuming do thermal stress concentrations in the wire i.e. tight bends, conduit etc...

I assume you are only interested in thermal management and not mechanical strength of the wire under load.

 

[Phrak]

Degradation of insulation

On voltage drop: it is a concern, but not a great one, since distances are relatively short (between 1 and 3 meters).

In this case your remaining concerns are degradation of insulation in your environment of interest, and degradation/heating of what the insulation contacts. I assume the conductor is copper, so that issues with aluminum don't come up.

You don't want your insulating material plasticizing so that the conductor migrates out of one side under strain, or undergoing slow oxidation, or melting, even.