Leak Rate Calculation

[PaytonA]

Could someone tell me how the ResMed machines calculate the leak rate? Is it something like the following?

Calculate the volume of a breath by the area under the air flow graph during inhale or exhale.

Multiply by breaths per minute to get breathing volume per minute.

Add the vent flow rate.

Subtract from total flow to get leak.

I am interested to know if this is correct.

Thanks,

PaytonA

[robysue]

I don't know for sure, but I think it's a simpler process than that. Among other things, I don't think the machine samples the leak flow data as often as the back pressure data used to measure the wave flow for air flow into/out of the lungs.

The CPAP, mask, hose, and our upper airway form a semi-closed pressurized system. The machine adds air to the system by blowing air into the system and the machine can track the rate it is adding air into the system at the blower end. And I believe that leak data is really based very simply on the amount of air the machine adds to the system each minute in order to maintain the appropriate pressure setting.

Once the desired pressure setting is reached, if there were NO leaking at all, NO additional air would need to be blown into the system to maintain the pressure, once enough air had been added to the system to reach the desired pressure. Think about a properly inflated and non-leaking tire: Once you've added enough air to pressurize the tire, you don't need to keep adding air to keep the tire at the desired pressure.

Of course if there were NO leaks at all in our CPAP system, then we'd be breathing our own exhaled air, and that would be very, very bad. So CPAP masks have in a built-in "leak rate"---they are designed to leak air out of the system at a known rate. In other words, in order to maintain the desired pressure, the CPAP machine has to blow enough air into the semi closed system to compensate for the air lost through the exhaust vents in the mask (intentional leak rate) plus any air lost through other leaks (unintentional mask leaks and mouth leaks).

So once the desired pressure is reached, the total leak rate equals the rate the machine is blowing air into the semi-closed system in order to maintain the desired pressure setting. And the machine can track the rate it is blowing air into the system in order to maintain the desired pressure at the blower end of the system. And I believe that's all that most CPAPs do to record their leak rates since, except for Resmeds, all other CPAPs report total leak rates.

But Resmeds report unintentional leaks rather than total leaks. On a Resmed machine, the mask setting tells the machine what the (approximate) intentional leak rate for the mask is based on the mask type selected. The approximate intentional leak rate probably includes some kind of a standard margin of error added since individual masks of the same make and model may have slightly different intentional leak rates in use. (Many mask's intentional leak rate tables report the intentional leak as **.* L/min +/- 5 L/min for example.) The approximate intentional leak rate for the selected mask type is then subtracted off the rate the machine is blowing air into the system (the unreported total leak rate), and the resulting number is reported as the (unintentional) leak rate.

An example may make this clearer.

I use an Swift FX nasal pillows mask at 8 cm H2O of pressure. The Swift FX user guide indicates that the intentional leak rate at 8cm of pressure is about 29 L/min. That means that if a CPAP machine is set to 8 cm H2O, the machine will need to add about 29 liters of air per minute to the system to maintain the 8 cm H2O of pressure.

So ...

As long as the machine is blowing air into the system at a rate of 29 L/min (or less), there is no unintentional leaking.

If I'm using a CPAP other than a Resmed, my CPAP will report the actual air flow through the system as the total leak rate and it's up to me (the user) to know that 29 L/min is a very desirable leak rate for my mask at my pressure and that with a leak rate of 29 L/min, there is no significant unintentional leaking going on.

If I'm using a Resmed machine, however, the mask setting tells the machine what to subtract off the rate it is blowing air into the system to calculate the reported leak. If I've correctly selected the "nasal pillows" setting, that will tell the machine to subtract off (about) 29 L/min from the airflow and record that number as the reported "leak rate". So if there's no unintentional leaking going on, the actual airflow into the system is (about) 29 L/min and the Resmed subtracts off 29 L/min off the airflow and reports a leak of 0.0 L/min. (If the actual airflow into the system is less than 29 L/min, then the Resmed "leak rate" is (technically) negative, but Resmed will the "leak rate" as 0.0.)

Now if an unintentional leak occurs, then the machine has to blow additional air into the system just to maintain the desired pressure setting. So, to continue with our example, let's suppose I do something that triggers a small leak during the night for me and the total airflow into the system goes from 29 L/min (the mask's intentional leak rate) up to 40 L/min just to maintain the required 8 cm of pressure.

On any machine other than a Resmed, the total leaks are now reported as 40 L/min---the amount of air the machine is blowing through the system in order to maintain the required 8 cm of pressure. And it's up to me (the user) to know that my mask's intentional leak rate is 29 L/min and that a leak of 40 L/min indicates that I've got an unintentional leak of about 11 L/min, which is a moderate sized leak that is NOT enough to worry about, unless it woke me up.

On a Resmed machine, however, the reported leak is just the excessive or [i]unintentional[i] leak. The Resmed machine takes the 40 L/min of air being pumped into the system to maintain pressure and subtracts off the expected 29 L/min intentional leak from my Swift FX at my pressure of 8cm and reports a leak of 11 L/min = 40-29 L/min. And since 11 L/min is well under the Red Line at 24 L/min, it's clear from the ResScan software that this leak is not big enough to worry about---unless it woke me up.

[herbm]

Robysue, I think you are correct but I do don't "know" for certain.

It just measures the input air required to keep the system "at pressure" and subtracts the EXPECTED leak for the mask type that is set.

It really doesn't know the mask leak rate but has some sort of algorithm to yield expected leak at various pressures based on mask type.

[robysue]

Calculate the volume of a breath by the area under the air flow graph during inhale or exhale.

The volume of one inhalation is the tidal volume, if I recall correctly. Tidal volume is usually measured in milliliters. Tidal volume is usually measured in liters per min (L/min). A typical tidal volume for an awake adult at rest varies quite a bit (based on size) but is usually in the neighborhood of 500 ml Note that 1000 ml = 1 L, so the average awake, but resting tidal volume is around 0.5 L. Tidal volume falls during sleep, often by 15% or more, particularly in REM.

Multiply by breaths per minute to get breathing volume per minute.

Tidal volume times breaths per minute is equal to minute ventilation.

At rest and while awake, a typical healthy adult's minute ventilation is usually between 5 and 8 L/min. During sleep, the minute ventilation in a healthy adult without SDB usually decreases by 15% or more (and sometimes quite a bit more).

So even at low pressures, the total leak rate is not likely to be affected by variations caused by individual breaths in a statistically significant way. Hence it would make sense for a company designing a CPAP to keep the algorithm for calculating leaks as simple as possible, and total airflow into the system needed to maintain pressure is a statistically reasonable way to define total leaks for the semi-closed pressurized system.

[PaytonA]

This makes some sense to me except for one thing. The machine (mine is a ResMed) must supply enough air to satisfy 3 things. 1-Air for the vent. 2-air for the breath. 3-air for the leak. The machine knows how much air (at least approximately) is required for the vent. That leaves 2 unknowns unless the machine is able to detect the flow at a point where neither inhalation nor exhalation is occurring. Thinking about this as I am writing, I suppose that it is possible that the machine could detect a flow that had only 2 components - vent flow and leak - so that it could easily establish the leak rate by subtracting out the vent flow and as you indicated it would be even easier to give the total leak rate (intentional plus unintentional leak).

Thanks for the info. I had not looked at it that way and as you can see from my discourse it took me a little while to get there.

Thanks again.

PaytonA

[herbm]

It "knows" (or can know) the inhalation from the exhalation (EPR is based on this) so it can presume the volume is the average of when these two start (or stop).

In this way it can allow for the expansion and contraction of the lungs.

[PaytonA]

Calculate the volume of a breath by the area under the air flow graph during inhale or exhale.

The volume of one inhalation is the tidal volume, if I recall correctly. Tidal volume is usually measured in milliliters. Tidal volume is usually measured in liters per min (L/min). A typical tidal volume for an awake adult at rest varies quite a bit (based on size) but is usually in the neighborhood of 500 ml Note that 1000 ml = 1 L, so the average awake, but resting tidal volume is around 0.5 L. Tidal volume falls during sleep, often by 15% or more, particularly in REM.

Multiply by breaths per minute to get breathing volume per minute.

Tidal volume times breaths per minute is equal to minute ventilation.

At rest and while awake, a typical healthy adult's minute ventilation is usually between 5 and 8 L/min. During sleep, the minute ventilation in a healthy adult without SDB usually decreases by 15% or more (and sometimes quite a bit more).

So even at low pressures, the total leak rate is not likely to be affected by variations caused by individual breaths in a statistically significant way. Hence it would make sense for a company designing a CPAP to keep the algorithm for calculating leaks as simple as possible, and total airflow into the system needed to maintain pressure is a statistically reasonable way to define total leaks for the semi-closed pressurized system.

Simplicity is to be strived for. With the tidal volume and the breathing rate known, the flow required per minute for inhalation is known. The flow required by the vent is known so any total flow more than that must be leak. My original quandary was how to obtain the tidal volume and I figured that would be obtained by determining the area under the flow curve during an inhalation. I guess the question I should have asked was how the tidal volume is determined.

Best Regards,

PaytonA

[zonk]

ResMed machines report unintentional leak
No need to get the calculator out, the machine does the calculation but you need to tell the machine (S9) the type of the mask using

ResMed Mask/Device Compatibility List
http://www.resmed.com/assets/documents/s...ow_eng.pdf


unintentional leak (from mouth, mask, etc) = total leak minus intentional leak
intentional leak = exhalation ports leak rate at various pressure setting

PR machines report total leak
total leak = intentional leak plus unintentional leak

[robysue]

FOLKS,

I had a major (and stupid) unit conversions problem in this post. I've taken the time to correct it here.

This makes some sense to me except for one thing. The machine (mine is a ResMed) must supply enough air to satisfy 3 things. 1-Air for the vent. 2-air for the breath. 3-air for the leak.

Air for the vent and air for the leak are the same thing. The machine does NOT know where or why air is leaking out of the system. And it is irrelevant (from the blower's point of view) if the system needs 35 L/min of air added to maintain pressure because:

• the mask's intentional leak (vent) rate is 30 L/min and there's a small leak in the cushion of 5 L/min
  
• the mask's intentional leak (vent) rate is 35 L/min
  
• the mask's intentional leak (vent) rate is 25 L/min and there's a modest leak through the corner of the mouth of 10 L/min
  

In all three cases, the machine has to blow into the system a grand total of 35 L/min to maintain the desired pressure.

Only the RESMED machines bother to try to distinguish between whether all the leak is intentional (and so the reported leak is 0.0) or whether some of the leak is unintentional. And since all you do is set a generic "mask type" on the Resmed, all the machine is doing is subtracting off a fixed number for the expected leak rate for that mask type at that particular pressure. But individual masks of the same make and model have (slightly) different leak rates when used by real PAPers night after night, and so regardless of what the Resmed machines use for the intentional leak rate for the mask, it is an approximation of the real intentional leak rate for that particular user on that particular machine when the mask is actually in use. When you use a PR machine you get a feel for what your intentional leak rate actually runs, and it can be as much as 5 L/min below or above the mask's advertised "intentional leak rate" at your pressure, but over time you do notice that your baseline leak rate usually stays within a 5 L/min range that is "normal" for you and your mask. (For example, my leak rate is almost always about 7 L/min what the charts say it should be. It's always been that way and I think it's just normal for me and my machine.)

As for "air for the breath:" The critical idea is is that our inhalations and our exhalations are roughly the same size. And the time needed to exchange the air is really not that important when the machine is calculating the leak rate for a one minute period. The leak rate graph is NOT the instantaneous leak rate, it is an average leak rate and as such, the data gathered for it is at intervals that are much less frequent than that needed for the wave flow.

And relative to the amount of air the machine is blowing into the system to keep it pressurized, the size of our individual inhalations and exhalations is not that much.

(MUCH irrelevant detail ommited)

On a minute-by-minute basis: We typically have a minute ventilation of 5-8 L/min. Hence there's already way more air being blown into the system (at least 20-30 L/min) that we can possibly ever breath into our lungs over the course of the minute: So most of the air the CPAP has to blow into the system is actually used to compensate for the (intentional) leaks built into the system to make sure that the air we exhale into the system is vented out of the system instead of being rebreathed.

And all the air we inhale from the system is (almost immediately) replaced by the same amount of exhaled air, which is immediately vented out of the mask's exhalation vents.

In other words:

During one minute, the machine blows enough (new) air into the system to maintain pressure. Since all the air in the system is replaced each minute we can assume that of the (new) air that's blown in, we inhale some of it, but all of the rest of it is vented through the leaks (both intentional and unintentional). Over the course of a minute, the air we inhale is replaced by the air we exhale, and so we get the following calculation concerning the total leak rate and the total amount of air pumped into the system by the blower unit:

• Total (new) air pumped into the system in one minute in order to maintain pressure
  = (new air we inhale in one minute) + (new air we don't inhale during that minute)
  = (air we exhale in one minute) + (new air that is vented or leaked during that minute)
  = total amount of air leaked out of the system during one minute.

In other words, if we make two modest assumptions:

• 1) The total amount of air we exhale during one minute is equal to the air we inhale in one minute
  2) The machine blows enough air into the system to replace all (or virtually) all of the air in the hose and mask (i.e. the machine is blowing at least 30 L/min)

then the fact that our exhaled air is immediately vented out of the system is enough for us to conclude that total leak rate and the rate the machine adds air to the system are equal to each other.

The machine knows how much air (at least approximately) is required for the vent.

No. The typical CPAP doesn't have a mask setting and has no idea how much air is needed for the intentional venting for the mask. Only the Resmeds have a mask setting and the mask setting simply tells the machine what number to subtract from the total leak rate to report as "leak".

That leaves 2 unknowns unless the machine is able to detect the flow at a point where neither inhalation nor exhalation is occurring.

Detecting inhalations and exhalations is actually done through measuring very subtle differences in "back pressure" at the blower's end of the hose. When you inhale the back pressure is reduced very slightly; when you exhale the back pressure is increased very slightly. These differences in back pressure are not enough to affect the rate the machine is adding air into the system, but they are enough to detect how much air is going into or out of the lungs---i.e. the wave flow. Detecting when NO air is flowing into or out of the lungs is critically important for detecting apneas.