[amluto]

Let's try the math for real.

The site claims that the device can provide 20 cm H₂O (which is ~1961 Pa). Apparently some hypothetical normal man breathes at a rate of 6 liters (i.e. 0.006 cubic meters) per minute at rest [1]. If you pump a volume V of fluid across a pressure difference of ΔP, you've done work V·ΔP, which is 11.76 Joules per minute or 0.1961 Watts. Keep in mind that this is the work done, not the power consumed, but let's be generous and assume a 100% efficient blower. We'll also assume that no power is harvested during exhalation.

In eight hours, that's 1.56 Watt hours.

Duracell's zinc-air chemistry claims up to 442 Watt hours / kg [2]. That's about 3.5 grams of zinc-air goo per night, ignoring packaging, blower loss, etc.

This seems entirely plausible to me.

[1] http://www.normalbreathing.com

[2] http://media.ww2.duracell.com/media/en-US/pdf/gtcl/Technical...

Edit: In the interest of precision, I should add that the analysis is accurate for an incompressible flow. Air is definitely compressible, so, if you pump air, more volume is going into the pump than is coming out, and you're heating up the air (adiabatically if your pump is any good). So, with a compressible fluid, you have to consider not on the work done moving the fluid across a pressure difference but also the energy you've stored in the fluid by pressurizing it. The pressure difference for CAP is tiny here, though, so this would be a miniscule correction to the math above.

[Alkim]

Well now I guess I have to do the math:

W = P(b)V(b) ln(P(a)/P(b))

(https://en.wikipedia.org/wiki/Compressed_air_energy_storage)

CPAP requires roughly 10 cm of water pressure = 0.01 Bar Sea level atmospheric pressure is roughly 1 bar So the device has to pressurize to 1.01 Bar Now I’ve looked it up, and a typical CPAP machine moves 20 to 60 liters per minute Let’s say 30 liters per minute = 1800 liters per hour

W = (1.007)(1.8)ln(1/1.007) = (1.81)*(-00698) = 0.0127 MJ = 3.53 Watt Hour

So assuming perfect efficiency you would still need 28.64 watt hours of battery, just to do the compression.

470 Wh/kg for Zinc-air battery implies this device requires a 60gm battery (2 oz to those of us of a certain age)

I looked up CPU fan efficiency (what they say they derived it from) and PEAK fan efficiency only approaches 25%

http://www.nmbtc.com/fans/white-papers/fan_efficiency_import...

So, you’d really need an 8 oz battery hanging from your nose to make this work, not counting what you need to run the electronics, etc.

[amluto]

Where does that 30 liters per minute go? I'm reasonably sure it's not going into the user's nose or mouth.

Part of the claimed innovation here is that the thing sticks directly into the user's nose. If it seals well, it's plausible that there's very little wasted air, which would account for the factor of roughly five.

(Also, you used the formula for isothermal compression, I think. A tiny little device like they showed would probably move air quickly enough through its fan that it would be closer to adiabatic. This doesn't matter much for the small pressure change involved.)

[dragonwriter]

> Part of the claimed innovation here is that the thing sticks directly into the user's nose.

Putting the whole machine in the nose is novel, but I don't see how that is innovative in terms of changing required functionality compared to nose-only masks, which are already a thing.

[amluto]

It shouldn't have anything to do with the required technology. It might allow a considerable reduction in leakage, though.