Pressure, Flow, and Leaks

[chill]

I've seen a lot of posts that say these CPAP machines increase pressure in response to leaks. That does not make sense to the part of my brain that remembers a bit of my university physics. If there is a leak pressure should drop, the same as punctured tire. The machine would increase flow (volume of air) to compensate for the leak and maintain therapy pressure (up to the point where it can't provide more flow which I think is the "large leak" point). It could overshoot and provide more additional flow than what is leaking which would increase pressure, at least temporarily. The machine increasing pressure in response to anything not apnea related seems wrong.

Am I thinking about this correctly?

Chuck

[zonk]

I think pressure rise due to leaks is counter productive and can make leaks even worse
see post #2 ... Philips Respironics reply to the first question
http://www.apneaboard.com/forums/Thread-...743#pid743
The algorithm decreases CPAP pressure when the patient is experiencing a large mask leak

[DeepBreathing]

I may not be thinking quite straight today... However, as I understand it a CPAP machine has one and one only method of adjusting flow or pressure - and that is to vary the speed of the motor. If resistance remains constant (no leak) then increased motor speed will increase the pressure. If resistance is reduced (leak) then increased motor speed will increase the flow rate. I suspect that if a leak is detected the motor speed will increase to increase the flow rate, thus maintaining pressure, but only up to a certain limit, at which point the increase becomes counter-productive. (Which I think is more or less what Zonk said).

[FrankNichols]

I may not be thinking quite straight today... However, as I understand it a CPAP machine has one and one only method of adjusting flow or pressure - and that is to vary the speed of the motor. If resistance remains constant (no leak) then increased motor speed will increase the pressure. If resistance is reduced (leak) then increased motor speed will increase the flow rate. I suspect that if a leak is detected the motor speed will increase to increase the flow rate, thus maintaining pressure, but only up to a certain limit, at which point the increase becomes counter-productive. (Which I think is more or less what Zonk said).

Pretty much what I thought/think. The hose, airways and lungs represent a leaky column of air. The motor/turbine runs at a speed which adds air to the column which compensates for the loss of air do to leaks.

Statically (when not breathing) flow is equal to leak rate (if there were no leaks, there would be no flow). With constant leak rates (which they aren't - they change with changes in pressure - but we can pretend leaks are constant rate to simplify this) flow rate changes only compensate for increases or decreases in volume of the air column (breathing) and are temporary until the target pressure is reached. This maintains a static pressure in the column that is sensed at the sensor in the machine.

The air is a compressible medium and pressure changes at any location in the column propagate through the column in all directions at the speed of sound in that medium.

When we inhale there is a negative pressure wave that starts at the diaphragm and moves up the airways, through the hose to the sensor in the machine, the sensor senses the lowered pressure and increases the speed of the motor to add air to the column and maintain the target pressure. When we exhale the pressure at the diaphragm increases and a positive pressure wave travels up the column to the sensor and the motor its slowed to compensate and maintain/return the column to its target pressure.

When we breath, the column of air changes volume (the diaphragm moves) This results in a pressure waves traveling at the speed of sound from the lungs to the machine, which with the short length of the column - hose, airways and lungs, means the pressure wave moves from the diaphragm to the sensor almost instantaneously. (It is actually around 25 milliseconds if the column is 8 feet long - but will vary a little based on temperature, humidity, etc.) This also means changes in the speed of the motor/turbine are reflected in pressure changes at the diaphragm within 25 ms.

Flow rate is whatever the flow rate needs (up to the capacity of the turbine) to be to maintain the pressure. I don't know the rate of change in flow rate the turbine is capable of, but we do know it can create pulses are 4 hz, which is four cycles of 1 cm change in pressure per second - or 250 ms per cycle - in ResMed this is used to detect airway patency (open or closed state - if the airway is closed the pressure wave will be reflected back to the machine by the blockage).

If there were no leaks in the system, there would be no flow (flow rate would be zero) except during inhale and exhale. If the motor were a leakless compressor, it would not run except during inhale and exhale (it would run backwards to reduce pressure during exhaling). However, there are leaks, both in the turbine and in the vents at the mask (and some of our mouths ) so the machine constantly runs and there is a constant flow rate equal to the sum of the intentional (vents, turbine, etc) and unintentional (seals, mouth, etc) leaks.

I have read (meaning I don't know what they do) that machines increase pressure (flow) in response to leaks until a large leak is detected, then they (some of them?) reduce pressure/flow. The idea that the large leak might be a mask seal blow out and reducing pressure might allow the mask cushion/seal to reseal - from my personal experience, I expect my seals never reseal and I have to wake up and fix it.

I have no idea is any of this is anything other than a bad dream on my part, but that is my uninformed understanding.

[justMongo]

There are two definitions of flow that one must consider. There is the absolute flow from the generator. And, there is the flow in/out of the person's airway.

If we only consider the former, then it is counter-intuitive for the pressure to rise in response to leak. The machine simply increases flow to maintain constant pressure. The machine is also trying to figure out what the flow is to/from the person -- that periodic waveform and where to put the zero line such that inhalation volume = exhalation volume.

If the leak is small, then all is well and the machine just compensates. If the leak is huge, then pressure drops. (If you have autostart enabled, the machine may think you've removed the mask; and it will shutdown.) There is apparently a middle ground where the leak causes the machine to be unable to properly detect the flow to/from the person. Thinking the flow has diminished, the machine raises pressure.

That's my story; and I'm stickin' to it.

[Sleepster]

I've seen a lot of posts that say these CPAP machines increase pressure in response to leaks.

Flow rate increases in response to leaks. A lot of people don't form in their minds the distinction between flow rate and pressure that engineering types do. When I make air flow into a tire the pressure increases. Attempts at in-depth explanations of why analogies like this don't apply in other situations sort of bounce off the receiver, as if there's an impedance mismatch. Mathematics forms the basis of the understanding of the distinction between pressure and flow rate, and to say that most people are not good at math is an understatement.

We know that a lot of people are not good at algebra, but a lot of people are not good at arithmetical reasoning. They cannot, for example, do ratio reasoning. An example would be that the value of a ratio tends to decrease as the denominator increases. Or that the value of a ratio stays the same as long as the numerator and denominator change in direct proportion to each other. You can get fairly well into a discussion of this type with someone who will at some point ask a question like "The denominator is the one in the bottom of the fraction, right?"

They just don't see the value in understanding these things. They're not dumb, they're just different.

[chill]

Thank you to everyone for the confirmation and a special thanks to Frank for totally geeking out on it. There is a lot of technical depth on this forum, that is good to see.

[pholynyk]

Minor detail for the pedantic:

"(It is actually around 25 milliseconds if the column is 8 feet long "

The speed of sound in air at sea level is roughly 1126 feet/second, depending on density and humidity, as you said. As a sloppy engineer, I round that to 1000 ft/sec, or one foot per millisecond, which is easy to remember.

So it would be about 8 milliseconds for the 8 foot long column...

But the rest is all good.

[FrankNichols]

Minor detail for the pedantic:

"(It is actually around 25 milliseconds if the column is 8 feet long "

The speed of sound in air at sea level is roughly 1126 feet/second, depending on density and humidity, as you said. As a sloppy engineer, I round that to 1000 ft/sec, or one foot per millisecond, which is easy to remember.

So it would be about 8 milliseconds for the 8 foot long column...

But the rest is all good.

Thank you, of course you are right, I don't recall why I said 25ms - oh well.

[justMongo]

...I don't recall why I said 25ms - oh well.

Working in SI units?