I'm looking at this data sheet for this MeanWell power supply: (5v 100A)
https://www.jameco.com/Jameco/Products/ProdDS/2337471.pdf

3 Questions:

1. The typical input power of 12A / 115v is 1380 watts. It states an efficiency of 86%. Since the end output is 5v 100A (500 watts), I would expect the input wattage to be somewhere around 580-590 watts. Why is the input wattage so much higher than the output wattage? The gap is much larger than the efficiency losses. I ask because I will need a lot of these for a project, somewhere around 120-160 of these! So trying to ballpark how many circuits on a the breaker box I will need to supply all of these. Since the output 500-600 watts, I was hoping I could power 3x of them per a typical 15Amp breaker (1800w). But the large input amperage/wattage seems to indicate I can really only have 1 or 2 per circuit unfortunately.

2. To further complicate the input amperage/breaker/circuit count, the inrush current is quite high at 35A. I assume inrush current only sustains for a few milliseconds. Do "normal" circuit breakers account for such a high inrush current, or will special breakers be needed?

3. Is there any way to estimate the power usage of these power supplies when no power is being drawn from them? Are these types of power supplies pretty efficient at idle?

Thank you!

 

1. The typical input power of 12A / 115v is 1380 watts. It states an efficiency of 86%. Since the end output is 5v 100A (500 watts), I would expect the input wattage to be somewhere around 580-590 watts. Why is the input wattage so much higher than the output wattage?

It's likely not that much power.
The reason could be a poor AC input power factor, which would give a higher input VA (VAR) rating than the actual input power.
But the breakers would need to handle the VAR current rating of the supplies.

2. To further complicate the input amperage/breaker/circuit count, the inrush current is quite high at 35A. I assume inrush current only sustains for a few milliseconds. Do "normal" circuit breakers account for such a high inrush current, or will special breakers be needed?

Likely a standard breaker would handle that inrush.

3. Is there any way to estimate the power usage of these power supplies when no power is being drawn from them? Are these types of power supplies pretty efficient at idle?

I would think it would be less than 10% percent of the rated power, but the current may be higher, again due to poor power factor.

 

A low-cost power meter such as the Kill-A-Watt can be used to measure the actual input power, power-factor, and current of the power supplies.

 

I think that typical current draw figure quoted on the datasheet is possibly in error – I’ve seen this on a few Meanwell datasheets where they are using the same format; and it appears that someone has carried forward values from a datasheet they are modifying.

The attached Meanwell datasheet gives a 115V input current of 5.6A for a 5V 100A PSU, which is around the value I'd expect.

https://www.meanwell.com/webapp/product/search.aspx?prod=RSP-750

 

Since it is unlikely that all will be loaded at 100%, you should be able to run 3 off a 15A breaker. If you are going to run around 160 of these supplies, I’d consider adding a power up sequencer to minimise the inrush current causing nuisance tripping. You might even want to investigate a bespoke solution.

The link below is designed for UPS, but you can get them for general power up switching to limit the inrush current.

https://www.netio-products.com/en/glossary/powerup-sequence

 

I think that typical current draw figure quoted on the datasheet is possibly in error

i have doubts about that.
The device you referenced has power factor correction and lists the low power factor value, whereas the one referenced by the TS does not, specifically stating:

meaning it is not power-factor corrected.

So it might be better for the TS to use one like you referenced with power-factor correction built-in.

 

I’ve never seen a power factor that bad on a PSU, and Meanwell are quite a good brand.

If the power factor really is that bad, it would be worth adding external power factor correction, if powering 160 of them.