[bobbobobbo]

I understand a circuit should not exceed 80% continuous power. So for a 20A 120v circuit thats 1920watts.

I am looking to power some servers and would be running around ~2000 watts per cluster continuous. (85%)

(I work alongside as an apprentice so I'm still learning) I've never gone to this level of power so I'm curious what would occur. Will it work, and at worst would the breaker trip after a period of time? The current wiring in place was installed in 2007.

 

[Dennis Alwon]

A circuit may be 100% of the capacity as long as it is not continuous- 3 hrs or more.

 

[Dennis Alwon]

I understand a circuit should not exceed 80% continuous power. So for a 20A 120v circuit thats 1920watts.

I am looking to power some servers and would be running around ~2000 watts per cluster continuous. (85%)

(I work alongside as an apprentice so I'm still learning) I've never gone to this level of power so I'm curious what would occur. Will it work, and at worst would the breaker trip after a period of time? The current wiring in place was installed in 2007.

If it is continuous then it cannot be more than 80% unless the breaker is rated 100% for continuous loads. I doubt you will find one that small.

 

[bobbobobbo]

A circuit may be 100% of the capacity as long as it is not continuous- 3 hrs or more.

Yes yes I understand that. I am inquiring regarding continuous 24/7 power use. Let's say at an 85% rating. What would occur?

 

[Pete m.]

Yes yes I understand that. I am inquiring regarding continuous 24/7 power use. Let's say at an 85% rating. What would occur?

Probably nothing short of a "red tag" from your inspector. In other words it is not code compliant.

Pete

 

[Dennis Alwon]

Yes yes I understand that. I am inquiring regarding continuous 24/7 power use. Let's say at an 85% rating. What would occur?

The breaker is not designed for 100% so at 85% it will probably break down sooner.

 

[Big John]

Yes yes I understand that. I am inquiring regarding continuous 24/7 power use. Let's say at an 85% rating. What would occur?

The breaker may trip.

Per UL489, molded case breakers are tested for temperature rise during full load. If the temperature never plateaus, or they exceed a maximum threshold, they are not suitable to carry a continuous load because the overheating may cause the thermal element to operate. But if they do pass that they are listed for 100% continuous load.

Breakers that don't pass are then re-tested to be sure that they can withstand an 80% load with no adverse temperature rise, and those are the majority of breakers listed.

 

[kbsparky]

Personally, I would not design an installation with circuit(s) loaded beyond the 80%. Especially for servers. :blink:

What would eventually happen is the breaker(s) could trip out, due to continuous heating and loading, shutting down those precious servers when you need `em the most.

Not a pretty sight. :no:

 

[gnuuser]

at 85 percent the breaker would be generating more heat than it was nominally designed for
it will eventually trip if the temp gets too high and that includes when the ambient temp raised too high
this nuisance tripping may cost you more in the lost data, early failure of ups batteries and seriously shortening the life of the breaker

considering this info install the proper sized conductors and a sufficient breaker for the task

 

[bobbobobbo]

Thank you all for the very informative comments.

This was all mostly just conceptual, I was more curious what would realistically happen. I suppose that extra 5% makes a difference. Which makes me cautious of even loading a continuos source up to 80%..

 

[don_resqcapt19]

All breakers are designed and tested to operate at 100%, but not in an enclosure with other breakers. The 20% derating is because of the heat from adjacent breakers. It would not be uncommon for a breaker in an enclosure to hold 90% or more forever. The standard permits them to hold 134% forever and they must trip within in an hour or less at 135%.

That being said the code says you can't run them over 80%.

 

[FlyingSparks]

In each IDF room at the facility I work in there is a 20amp receptacle (single not duplex) for each switch. The network switches only use 160w each, or about 8.3% of that 80% continuous rating. The network switches are wired so they fail over, the concern in doing this is not for load, but for short circuiting. Having them all on the same circuit would prevent the designed fail over network mechanism. Depending on the facility's usage you should think of fail over protection in cases of short circuits.

 

[RIVETER]

I understand a circuit should not exceed 80% continuous power. So for a 20A 120v circuit thats 1920watts.

I am looking to power some servers and would be running around ~2000 watts per cluster continuous. (85%)

(I work alongside as an apprentice so I'm still learning) I've never gone to this level of power so I'm curious what would occur. Will it work, and at worst would the breaker trip after a period of time? The current wiring in place was installed in 2007.

Historically, copper wiring is very forgiving. It is a great conductor. However, somewhere along the line limits have to be set and wire Gage and the insulation surrounding it are taken into account. Overcurrent...within reason does not normally affect the copper because it is so "FORGIVING", but, over time an excessive current and the subsequent heat can possibly bake the insulation.

 

[Meadow]

The breaker may trip.

Per UL489, molded case breakers are tested for temperature rise during full load. If the temperature never plateaus, or they exceed a maximum threshold, they are not suitable to carry a continuous load because the overheating may cause the thermal element to operate. But if they do pass that they are listed for 100% continuous load.

Breakers that don't pass are then re-tested to be sure that they can withstand an 80% load with no adverse temperature rise, and those are the majority of breakers listed.

Chances are it will hold. The 80% rule comes from fully loaded panels where heat build up takes place, increasing the chance of a nuisance trip. I see 100% loading often, but again its not code.

 

[Big John]

Chances are it will hold. The 80% rule comes from fully loaded panels where heat build up takes place, increasing the chance of a nuisance trip. I see 100% loading often, but again its not code.

As far as I know UL only looks at temperature rise on the breaker in a fixed ambient temperature. I don't believe they take neighboring breakers into consideration, but I'll dig up the text of the tests.

 

[RIVETER]

Chances are it will hold. The 80% rule comes from fully loaded panels where heat build up takes place, increasing the chance of a nuisance trip. I see 100% loading often, but again its not code.

So, are you saying that you have never heard of the excessive current...over time, causing a breakdown of the insulation of the wiring? Circuit breakers are not involved here.

 

[papaotis]

somebody help with the math please. 80% is 2000 watts -80 watts?:blink:thats how you would come up with1920

 

[FlyingSparks]

somebody help with the math please. 80% is 2000 watts -80 watts?:blink:thats how you would come up with1920

20a * 120v = 2400VA <-- Non-continuous load (0, 3)hr
2400VA * 0.80 = 1920VA <-- Adjusted for continuous [3, ∞)hr

Continuous Load. A load where the maximum current is expected to continue for 3 hours or more.

 

[RIVETER]

somebody help with the math please. 80% is 2000 watts -80 watts?:blink:thats how you would come up with1920

Are you on the right forum?

 

[papaotis]

thanks. i needed a reminder that different methods come up with different results! i simply took 2000x 80%:whistling2: