How an Anesthesia Machine Performs a Leak Test

How anesthesia machines pressurize the breathing system, watch for pressure loss, and use the result to detect leaks before clinical use

An anesthesia machine leak test is designed to answer a simple question:

Published August 16, 2026 · Revised September 6, 2026

Back to Biomed Basics

What This Page Explains

This page covers:

A leak test seals a defined part of the machine, pressurizes it or draws it below atmospheric pressure, and observes how pressure changes after flow stops. Stable pressure or a leak rate within the specified limit supports that the tested volume is sufficiently tight. Unexpected pressure change indicates gas entering or leaving that boundary.

The boundary is the key. A breathing-circuit test, low-pressure-system test, vaporizer test, and automated checkout may include different valves, hoses, canisters, seals, and modes. Passing one does not prove every gas path is leak-free, and using the wrong APL or bag/ventilator position can invalidate the result.

Worked Example: Automated Leak Test Fails

Reinstall the circuit and required test adapter exactly as the manufacturer specifies, confirm APL and bag/ventilator positions, seat the absorber canister and water traps, and inspect breathing hoses, sampling connections, vaporizers, and seals. Repeat once without changing several things at the same time.

If the failure remains, isolate approved sections using the service procedure and record the measured leak rate or pressure decay. Never use oxygen where the procedure specifies another gas, apply unapproved pressure, or release the machine until the complete required checkout passes.

What Is a Leak?

A leak is an unintended path that allows gas to escape from, or enter, a system.

Possible leak locations include:

Anesthesia machines depend on controlled:

A leak can affect:

Not All Leak Tests Test the Same Parts

This is one of the most important concepts.

A machine may have separate tests for:

A passing test in one area does not prove the entire machine is leak-free.

Know the Test Boundary

Before interpreting any result, ask:

What part of the machine was sealed and pressurized during this test?

That tells you what the result actually proves.

Positive-Pressure Leak Test

A positive-pressure test raises system pressure above atmospheric pressure.

Then the system is observed for:

System pressurized to:

30 cmH2O.

Fresh-gas flow stopped.

Pressure falls rapidly.

There is a leak somewhere in the tested volume.

Pressure Decay

If pressure drops over time, gas is leaving the system.

The faster the drop:

Generally, the larger the leak.

But the relationship depends on system volume and compliance.

Why Compliance Matters

Flexible components can:

after pressurization.

That can cause a small initial pressure change even in a relatively tight system.

Use the manufacturer-defined test method.

Negative-Pressure Leak Test

Some systems are tested by creating pressure below atmospheric pressure.

Conceptually:

This is especially useful for certain low-pressure gas-delivery systems.

Negative-Pressure Bulb

Traditional anesthesia-machine checks may use a suction bulb.

The bulb is squeezed to create negative pressure.

If it reinflates unexpectedly, air may be entering through a leak.

Positive vs Negative Testing

These can test different failure paths.

A component might seal better under:

than:

Negative pressure.

That is why the manufacturer specifies the correct procedure.

Automated Checkout

Modern anesthesia machines often perform automated leak testing during startup.

The machine may control:

and calculate a leak rate automatically.

Automated Leak Test Does More Than Watch a Gauge

The machine may:

Isolate a test volume.

Apply known pressure or flow.

Measure pressure response.

Calculate compliance.

Estimate leak.

Compare result to limits.

Leak Rate

Instead of simply saying:

some machines report leak as:

This estimates how much gas escapes per unit time.

Example

Measured system leak:

40 mL/min.

Allowed maximum:

100 mL/min.

Pass.

Measured:

300 mL/min.

Fail.

Use the actual manufacturer limit.

Leak Rate Is Not the Same as Pressure

A machine may pressurize to:

and then calculate how much flow is required to maintain that pressure.

That flow can be used to estimate leakage.

Pressure Sensor Matters

An automated leak test depends on accurate pressure sensing.

If the sensor is wrong, the machine may report:

If the test calculates leakage from flow, the flow sensor also matters.

A false high flow reading could make the machine believe the system is leaking more than it really is.

Leak Test Failure Does Not Automatically Mean Real Leak

First determine whether an independent test confirms the problem.

A sensing failure can produce a false leak-test result.

Breathing Circuit

The external breathing circuit is one of the easiest places to leak.

Check:

before opening the machine.

Circuit Connections

A connection can look attached but not be fully seated.

Possible symptom:

Machine repeatedly fails leak test after circuit change.

Cracked Hose

Small cracks may open only when the hose is:

Flex the hose visually during controlled testing.

Reservoir Bag

A damaged reservoir bag can leak significantly.

Check for:

The bag connection itself can leak.

Do not assume the bag is the only part.

APL Valve

The adjustable pressure-limiting valve controls pressure during manual or spontaneous ventilation.

During certain leak tests, it may need to be:

If the test requires the APL closed and it is left open:

Gas escapes intentionally.

The machine fails the leak test.

Nothing is actually broken.

APL Valve Leak

A valve that cannot fully close may cause:

The bag/ventilator selector changes which gas path is active.

Wrong position can include or exclude different parts of the system.

Manual Mode Test

A leak may appear only in:

Bag/manual mode.

That points toward components unique to that path.

Ventilator Mode Test

If leak appears only in ventilator mode, consider:

depending on design.

Inspiratory Valve

A damaged or poorly seated inspiratory valve can create abnormal flow paths.

Depending on machine design, it may contribute to a leak-test failure.

Expiratory Valve

The same is true for the expiratory valve.

Inspect:

as allowed by the manufacturer.

CO2 Absorber Canister

The absorber assembly is a very common leak location.

Possible causes include:

A small gasket problem can create a surprisingly large breathing-system leak.

Absorber Removed and Reinstalled

If the machine passed before absorber service and fails immediately afterward:

Start there.

The timing is valuable evidence.

Quick Isolation

If the design allows an approved bypass or alternate absorber configuration, this may help isolate the leak.

Use only manufacturer-supported procedures.

Vaporizer

Some leak tests include parts of the vaporizer or low-pressure gas system.

A vaporizer can leak at:

Some vaporizer leaks can exist even when concentration control is set to:

OFF.

The exact design matters.

Vaporizer Mount

A poorly seated vaporizer may create a leak at the mounting manifold.

Check:

If allowed by manufacturer procedure, isolating vaporizers can help determine whether the leak follows one unit.

Low-Pressure System

The low-pressure gas-delivery system may include:

Leaks here can alter delivered gas composition.

This is different from a breathing-circuit leak.

Fresh-Gas Flow

During some tests, fresh-gas flow is used to pressurize the system.

Once the system reaches target pressure, flow may be reduced or stopped.

Fresh-Gas Flow Can Hide a Leak

If fresh gas continues flowing:

Pressure may stay normal even though gas is escaping.

You may only be proving:

Incoming flow exceeds leak flow.

That is not the same as proving the circuit is tight.

Gas Sampling Connection

The gas analyzer removes gas from the breathing system.

Depending on design, sampling may affect leak testing.

Sample Line Disconnected

An open sampling port may create an intentional leak path.

The machine may fail the test.

Sample Return

If sampled gas is normally returned to the system, a disconnected return line may affect results.

Scavenging System

Scavenging is related to the breathing system but should not normally create uncontrolled patient-circuit suction or pressure.

A scavenging-interface problem can sometimes affect pressure behavior.

Do Not Confuse Scavenging Flow With Breathing-System Leak

They are separate systems connected through controlled interfaces.

Ventilator Bellows

Some anesthesia machines use bellows.

Leaks may occur at:

A damaged bellows can cause:

Some bellows rise during expiration.

A leak may cause them to:

This can be a useful visual clue.

Piston Ventilator

Piston systems may leak through:

The test method differs from bellows systems.

Turbine-Based Systems

Modern machines may use blower or turbine architectures.

Again, do not assume all anesthesia ventilators behave the same way.

Leak vs Compliance Test

Automated checkout may test both:

These are different.

Circuit Compliance

Compliance measures how much system volume changes with pressure.

A circuit can be:

but still:

Leak-free.

Leak and Compliance Can Look Similar Initially

Both can cause pressure to change during pressurization.

The machine's algorithm separates them using timing and known flow/pressure behavior.

Incorrect Circuit During Checkout

If the machine performs compliance compensation with one circuit and you replace it afterward, delivered-volume compensation may be wrong.

Complete checkout with the intended clinical setup where required.

Test Lung

Some checkout procedures require:

Use exactly the specified setup.

Y-Piece Occlusion

Many leak tests require the patient connection to be sealed.

If it is not:

The machine is open to atmosphere.

Leak test will fail immediately.

Thumb Occlusion

Some older manual procedures may have users occlude the Y-piece.

Modern manufacturer instructions may specify a test plug or other method.

Follow the approved method.

Leak Test Plug

A dedicated plug creates a repeatable seal.

If the plug itself is damaged:

You can create a false leak.

Pressure Gauge

Older machines may rely on a mechanical pressure gauge.

Verify whether the gauge itself is accurate if results do not make sense.

Electronic Pressure Display

Modern machines display pressure electronically.

Again:

Displayed pressure is only as good as the pressure sensor.

Independent Pressure Measurement

A calibrated anesthesia or ventilator analyzer can help confirm:

depending on test setup.

External Leak Tester

Some service procedures use dedicated:

Follow the specified equipment.

Soapy Water?

Bubble solutions can sometimes locate gas leaks in certain systems, but they are not universally appropriate for anesthesia equipment.

Do not apply liquids to internal components unless manufacturer procedures specifically permit it.

Listen for Leaks

A large leak may be audible.

This can help localize:

But:

No audible leak does not mean no leak exists.

Feel for Gas?

Do not rely on your hand as a quantitative leak detector.

Use the approved test method.

Isolation by Sections

A powerful troubleshooting approach is to reduce the test volume.

Conceptually:

Whole system leaks.

Then isolate:

until the leak disappears.

One Change at a Time

If you:

you do not know what fixed it.

Failure Follows Component

Example:

Machine fails with Vaporizer A.

Passes with Vaporizer A removed.

Vaporizer A fails on another compatible machine.

Failure follows vaporizer.

Strong evidence.

Failure Stays With Machine

Two known-good vaporizers:

Machine still fails.

Both work elsewhere.

Look at machine-side manifold or another common path.

Leak Only in Manual Mode

Possible areas include:

Possible areas include:

Mode comparison is very useful.

Leak Only With One Circuit

Known-good circuit:

Pass.

Original circuit:

Fail.

Failure follows circuit.

No internal repair needed.

Leak Only With Gas Module Attached

Machine passes without sample line.

Fails with sampling system connected.

Look at:

Machine passed before absorbent replacement.

Fails afterward.

Canister seal not seated correctly.

Timing helps isolate quickly.

Pressure Holds but Delivered Volume Is Low

A passing leak test does not prove ventilator delivery is accurate.

You still need:

Different tests answer different questions.

Leak Test Passes but Agent Odor Exists

Possible causes include:

The breathing circuit may be tight.

Leak Test Fails but Ventilator Seems Fine

A moderate leak may not be obvious during a short breathing test.

Do not dismiss the failed checkout.

Intermittent Leak

A seal may leak only when:

Reproduce the actual condition.

O-Ring Problems

O-rings can:

They are common leak sources.

Lubrication

Only use manufacturer-approved lubricants where specified.

Incorrect lubricant can damage materials or alter sealing.

Tightening Everything Is Not a Strategy

Overtightening plastic fittings can:

Use manufacturer procedures and torque requirements.

Automated Test Stops at a Specific Step

This can be valuable.

If checkout consistently fails during:

but passes:

the failure is already somewhat localized.

Read the Exact Error

A message such as:

provides more information than:

Checkout Failed.

Use the actual failure details.

Event or Service Logs

Some machines store checkout results.

Look for:

This can help with intermittent complaints.

Baseline Matters

If the machine historically measures:

and suddenly measures:

even though the official limit is:

the change may still be worth investigating depending on manufacturer guidance.

Do Not Invent Tighter Limits

Use the official acceptance criteria.

A trend can guide troubleshooting, but pass/fail should follow the manufacturer requirement.

Real-World Example: Failed Breathing-System Leak Test

Machine leak:

Large.

External circuit replaced:

Still fails.

Absorber canister reseated:

Pass.

Canister gasket was folded.

Real-World Example: Leak Only in Manual Mode

Ventilator mode:

Pass.

Bag mode:

Cannot hold pressure.

Reservoir bag has small tear near neck.

Real-World Example: Leak Only in Ventilator Mode

Manual bag holds pressure normally.

Ventilator mode fails leak test.

Bellows assembly has damaged seal.

Real-World Example: False Leak Result

Automated checkout reports excessive leak.

Independent pressure/leak test shows system tight.

Pressure-sensor calibration found offset.

The machine was measuring the test incorrectly.

Real-World Example: Vaporizer Leak

Machine passes with both vaporizers removed.

Fails when one particular vaporizer is installed.

Same vaporizer causes failure on another compatible machine.

Failure follows vaporizer.

Real-World Example: Gas Sampling Connection

Machine suddenly fails checkout after CO2 sample line replacement.

Sampling fitting not fully seated.

Once corrected, leak test passes.

Real-World Example: Fresh Gas Hides Leak

Manual pressure appears stable while fresh-gas flow remains high.

Fresh gas turned off according to test procedure.

Pressure rapidly falls.

The system was leaking all along.

Common Mistakes

Treating “Leak Test Failed” as a Diagnosis

It only tells you the tested system leaked beyond the limit.

Not Knowing What the Test Includes

Understand the test boundary.

Leaving the APL Valve Open During a Test That Requires It Closed

That creates an intentional leak path.

Replacing Internal Parts Before Checking the External Circuit

Start with easy components.

Ignoring the Absorber Canister

Seals there are common failure points.

Assuming a Passing Leak Test Proves Ventilator Volume Accuracy

It does not.

Letting Fresh Gas Flow Hide the Leak

Follow the exact procedure.

Changing Several Parts at Once

Follow the failure.

A Useful Troubleshooting Framework

For a failed anesthesia leak test, ask:

What exact test failed?

Then:

What parts of the machine are included in that test?

Then:

Is the external circuit known good and correctly sealed?

Then:

Does the failure change between manual and ventilator modes?

Then:

Does isolating the absorber, vaporizer, or sampling system change the result?

Then:

Does an independent pressure or leak measurement confirm the machine's result?

That turns:

into a manageable isolation process.

Another Useful Question

Ask:

Where can gas escape from the volume I just pressurized?

That is the physical question behind nearly every leak test.

What Did You Actually Prove?

If an anesthesia machine reports:

you proved:

The specific system volume tested met the manufacturer's leak criteria under that test condition.

You did not necessarily prove:

If the test fails, you proved:

The machine detected excessive pressure loss, flow requirement, or another leak-related condition in the tested system.

You still have to find the source.

Final Thoughts for Biomeds

Anesthesia-machine leak testing is easiest when you think about:

Test boundary → Pressure → Escape path.

First determine what the machine is actually testing.

Then ask:

Where can gas leave this sealed volume?

Start with:

Then move inward.

Compare:

Manual vs ventilator mode.

Remove or isolate components one at a time when the manufacturer allows it.

And remember:

A failed leak test is not the repair diagnosis.

It is evidence that:

Something inside the tested boundary is not holding gas the way it should.

Your job is to make the leak follow one component or one section of the machine.

— Jake

Important Note

Anesthesia-machine leak-test methods, tested boundaries, acceptable leak rates, pressure targets, automated checkout algorithms, vaporizer testing, ventilator isolation procedures, and service tools vary substantially by manufacturer and model. Follow current manufacturer checkout and service documentation exactly, use approved test equipment and configurations, and complete all required breathing-system, ventilator, gas-delivery, scavenging, alarm, and safety verification before returning the machine to clinical use.

Related Biomed Basics

Back to Biomed Basics