Refrigerant evacuation field guide: pulling a deep vacuum

Evacuation, and why it decides whether the system lives

Evacuation is pulling a deep vacuum on a sealed system to drag out the air, the moisture, and the other non-condensable gases that got in while the system was open, before you put any refrigerant in. A vacuum pump lowers the pressure inside the system far below atmospheric, low enough that water trapped in the lines boils away to vapor and gets carried out with the air. You do not charge a system you did not evacuate. That is the rule the whole job hangs on.

This is the make-or-break step of an install or a repair, and it is the one that gets shortchanged because nobody can see a vacuum. A good braze you can inspect. A clean flare you can feel. A deep vacuum is invisible, so the temptation is to run the pump for fifteen minutes, watch the manifold needle peg, and call it done. The system that gets that treatment runs for a while and then dies of something nobody connects back to the day it was installed.

The work is a chain: recover the old charge if there is one, pressure test for leaks with nitrogen, pull the vacuum, prove it held with a decay test, then charge. Each step protects the next. The leak detection and recovery guide covers the recovery and the leak test in depth, and the charging guide picks up where this one ends. This guide is the middle link, the vacuum itself, which is the step that either dries the system out or quietly leaves the seeds of its failure inside.

What air and moisture actually do inside the system

Moisture is the worst thing you can leave in a refrigeration system, and it does its damage three ways. It reacts with the refrigerant and the oil to form acids, and those acids eat the compressor windings and the bearings from the inside until the motor shorts or the mechanicals seize. It freezes at the metering device, where the refrigerant is at its coldest, and a slug of ice at the orifice or the TXV chokes the flow and the system loses capacity in a way that comes and goes as the ice forms and melts. And it sludges the oil, breaking down the lubricant the compressor depends on.

Air is the other half of the problem, and it shows up as non-condensables. Air will not condense at the pressures and temperatures the refrigerant works at, so it collects in the top of the condenser and takes up space the refrigerant needs. That raises head pressure, which makes the compressor work harder, run hotter, and pull more power to move less heat. The system loses capacity and the compressor's life gets shorter at the same time.

Put the two together and you have why the vacuum exists. The deep vacuum is the only practical way to get both out: it boils the water to vapor and pulls it, and it pulls the air with it. There is no additive that fixes a wet system the way the vacuum does, and a filter drier catches what is left, not a system soaked through. Skip the evacuation and you have built the failure in. It just takes a year or two to surface.

What is a micron, and how deep do you pull?

A micron is a unit of absolute pressure used to measure a deep vacuum, equal to one-thousandth of a millimeter of mercury. Atmospheric pressure is about 760,000 microns, so when you read 500 microns on the gauge you are down at a tiny fraction of normal pressure, almost a perfect emptiness inside the system. The lower the micron reading, the deeper the vacuum and the less air and moisture is left.

A common target is to pull the system to 500 microns or below before charging. That figure shows up across the trade and in a lot of manufacturer literature, but it is the equipment manufacturer's evacuation spec that governs, so confirm the number for the unit in front of you. Some systems and some oils call for going deeper. The point of the number is not the number itself. It is that a deep vacuum proves the air is out and gives the moisture a chance to boil away.

The thing to understand is that microns measure absolute pressure, counted up from a perfect vacuum, not gauge pressure counted from atmosphere. That is why the manifold gauge cannot read it. A manifold gauge measures pressure relative to the air around you, and its low-pressure end just shows the needle pinned somewhere near 30 inches of mercury long before you get anywhere near 500 microns. The micron is a different measurement on a different scale, and you need the right instrument to see it.

Micron

A unit of absolute pressure, one-thousandth of a millimeter of mercury; the lower the reading, the deeper the vacuum

Absolute pressure

Pressure measured up from a perfect vacuum, not from atmosphere; how a deep vacuum is read

Non-condensable

A gas such as air that will not condense in the system, raising head pressure until the vacuum removes it

The micron gauge, and where to put it

An electronic micron gauge is the only tool that reads a true deep vacuum, and it is not the same as the gauge on your manifold. The manifold's compound gauge tops out its vacuum scale at roughly 30 inches of mercury, and 30 inches of mercury still leaves you tens of thousands of microns from the target. The needle pegs and stops telling you anything useful while the real work of the evacuation is still ahead of you. The micron gauge uses a thermistor or a similar sensor that actually resolves the range from a few thousand microns down to single digits.

Where you put the gauge matters as much as having one. Place the micron gauge on the system, as far from the pump as you can get it, not at the pump itself. The pump is the deepest point in the setup, so a gauge mounted there reads a beautiful number that the system never sees. The restriction of the hoses and the cores means the system is always shallower than the pump while you are pulling. Read the gauge at the equipment, on the opposite side from the pump connection where you can manage it, and you read the vacuum that the refrigerant will actually live in.

Mount it with a core depressor or a tee at a service port, keep it out of any oil that could foul the sensor, and let it settle before you trust the number. A micron gauge reading at the pump is the single most common way a tech fools himself into thinking a system is evacuated when it is not.

Micron gauge

Electronic vacuum gauge with a thermistor sensor that reads deep vacuum in microns; the manifold gauge cannot

Compound gauge

The manifold's low-side gauge; its vacuum scale stops near 30 in Hg, far short of the micron range

The vacuum pump and the oil that runs it

Use a two-stage vacuum pump for refrigeration work. A two-stage pump runs two rotors in series, so the second stage starts where the first leaves off and the pump can pull down into the low microns a single-stage pump cannot reach. A single-stage pump is fine for pulling the bulk of the air, but it stalls out well above a real deep-vacuum target. The two-stage is what gets you to 500 microns and below.

CFM is the pump's displacement, how fast it moves gas, and it sets how quickly you pull down, not how deep. Most residential and light commercial work runs a pump in the 4 to 8 CFM range, and a bigger system or a long line set wants more CFM so the pull-down does not take all day. More CFM gets you to the target faster; the two-stage design is what lets you reach the target at all. Match the pump to the system in front of you.

Pump oil is the part that quietly kills more vacuums than any pump defect. Vacuum pump oil is what seals the pump and lets it reach a deep vacuum, and it absorbs the moisture you pull out of the system, which contaminates it fast. Oil that is milky, dark, or low will not let the pump pull down no matter how good the pump is. Change the oil before a big evacuation and again after pulling a wet system, run the pump with the ports capped and check that it can pull itself down to its rated blank-off vacuum, and you will catch a tired pump before it costs you an afternoon. A pump that cannot pull itself down to spec on clean oil will never pull a system down.

Two-stage pump

A vacuum pump with two rotors in series, needed to reach a deep vacuum below 500 microns

CFM

The pump's displacement, how fast it moves gas; sets pull-down speed, not the depth it can reach

Blank-off vacuum

The deepest vacuum a pump can pull with its ports capped; a check on the pump and its oil

Large hoses and pulling the cores

The number one reason a vacuum pulls down slow is restriction, and the restriction lives in the hoses and the Schrader cores. A standard 1/4 inch refrigeration hose is fine for reading pressures, but it is too small to move gas at vacuum, and at deep vacuum the gas is so thin that the small bore strangles the flow. The pump can be moving plenty of CFM at its inlet and almost nothing at the system, because the hose between them is the bottleneck.

Two changes fix most of it. Pull the vacuum through large-diameter hoses, 3/8 inch or 1/2 inch, and keep them short, because every foot of hose and every reduction in bore adds restriction. Then remove the Schrader valve cores with a core-removal tool. A Schrader core is a tiny spring-loaded valve, and at vacuum it is a major choke point. A core-removal tool lets you take the cores out while the system stays sealed, pull the vacuum through the wide-open port, and put the cores back at the end without breaking the vacuum or losing the charge.

This is the setup that separates a two-hour evacuation from a six-hour one on the same system. Crews that pull every vacuum through 1/4 inch hoses with the cores still in wonder why their evacuations drag and why their decay tests look wet, when the real problem is that the system never actually got deep, it just got slow.

Core-removal tool

A valve that lets you take out the Schrader cores while the system stays sealed, removing a major flow restriction

Evacuation hose

A large-bore hose, 3/8 in or 1/2 in, that moves gas at vacuum far faster than a 1/4 in refrigeration hose

How do you set up for a fast deep vacuum?

Set up to evacuate from both sides at once, through the biggest openings you can make, with the gauge reading where the refrigerant lives. The fast deep vacuum is a setup problem before it is a pump problem, and the setup is the same every time. Connect the vacuum pump to the system through large-diameter hoses, on both the high side and the low side, so you are pulling the whole system in parallel instead of dragging the vacuum through the metering device from one side.

Remove the Schrader cores at the ports you are pulling through, using core-removal tools, so the cores are not choking the flow. Put the micron gauge on the system at a port away from the pump, ideally on the opposite side, so it reads the vacuum the system actually reaches and not the deeper number at the pump. Keep the hose runs short and the connections tight, because a leak in your own setup reads exactly like a leak in the system and will fail your decay test for no reason.

Pulling from one side only, through the cores, with 1/4 inch hoses, is the slow setup that most people default to. It will eventually get there on a small dry system and it will never get there cleanly on a big or wet one. The few minutes it takes to pull the cores and hook both sides is paid back many times over in pull-down time, and more than that in a vacuum that is actually deep instead of just slow.

Pulling the vacuum

With both sides connected through large hoses, the cores out, and the gauge on the system, start the pump and let it run. The first stretch is the easy part, where the pump clears the bulk of the air and the micron reading falls fast, down through the thousands. Watch the rate of the fall, not just the number. A vacuum that drops quickly and keeps dropping is a tight, dry system. A vacuum that drops fast and then stalls is telling you something before you even get to the decay test.

Pull the system down past your target and give it margin, because the reading at the gauge while the pump runs is the system being held down by the pump, not the vacuum it will sit at once you isolate it. Getting to 500 microns with the pump running is not the same as the system holding 500 microns after the pump is valved off. The pull-down is where you reach the number. The decay test is where you find out if it means anything.

How long it takes depends on the system, the moisture in it, and the setup. A small, dry, properly rigged system can hit a deep vacuum in well under an hour. A large or wet system can run for hours, and trying to rush it is how you end up charging a system that never got dry. Let the pump do its work and judge it by the gauge, not the clock.

What is a decay test?

A decay test, also called a standing vacuum or rate-of-rise test, is how you prove the system is actually dry and tight after you reach the target vacuum. You pull the system down to target, valve off or isolate the vacuum pump so it can no longer hold the vacuum, and then watch the micron gauge. The pull-down gets you to the number. The decay test is the only thing that tells you the number will hold, and it is the step that separates a real evacuation from a guess.

Isolate the pump, not the system. Close a valve between the pump and the system, or use a core tool with a built-in valve, so the system is sealed off with the gauge still reading it. Then wait. How long depends on the system and the standard you hold to, but you want enough time for a problem to show, commonly on the order of ten to fifteen minutes or more, and longer on a large system. A tight, dry system holds its vacuum nearly flat. A wet or leaky one does not, and the way it rises tells you which.

Set the target and the decay criteria before you start, and write down what the gauge did. A vacuum that touched 500 microns for a moment and then climbed back to 5,000 was never evacuated, it was just briefly pulled down. The decay test is what makes the difference visible. Skip it and you are charging on hope.

Decay test

Isolating the pump and watching the micron gauge for a rise; proves the vacuum holds, not just that it was reached

Rate of rise

How fast the micron reading climbs after isolating the pump; its speed and shape separate a leak from moisture

Why is my vacuum rising after I shut off the pump?

A vacuum that rises after you isolate the pump is telling you one of two things, a leak or moisture, and the shape of the rise tells you which. Reading the decay correctly is the skill, because the fix is different for each. A leak you have to find and seal. Moisture you have to keep pulling.

A fast, steady climb that runs right back up toward atmosphere is a leak. Air is pulling in through a hole, and it will not stop, so the reading just keeps going up and does not level off. If the rise does not slow down and does not settle anywhere, stop, because you are not going to evacuate your way out of a leak. Go back to the nitrogen pressure test, find it, and fix it before you waste any more pump time.

A rise that climbs and then levels off and holds at some higher level is moisture still boiling off. The water in the system has a vapor pressure, and once the vacuum drops below that pressure the water boils and the vapor it gives off pushes the reading up until it equalizes, then it stalls there instead of running away. At normal room temperature that leveling-off tends to land somewhere up in the thousands of microns, which is a clear moisture signal. Keep pulling, give it time, and consider a nitrogen sweep. A flat hold, where the reading barely moves and stays low over the wait, is the one you want: the system is dry and tight, and you can charge it.

Why a deep vacuum boils the water out

The reason a vacuum dehydrates a system is that water boils at a lower temperature as the pressure drops. At atmospheric pressure water boils at 212 degrees F, but pull the pressure down into the deep-vacuum range and water boils at temperatures well below freezing. Down near 500 microns water turns to vapor at roughly minus 12 degrees F, which means the moisture in a system at ordinary shop temperature is being held above its boiling point by the vacuum. It boils, and the pump carries the vapor out.

This is dehydration, not just emptying. You are not sucking liquid water out of the lines, you are dropping the pressure low enough that the liquid water flashes to vapor and leaves as a gas. That is why the vacuum has to be deep and why it has to be held. A shallow vacuum never gets below the water's boiling point at that temperature, so the water just sits there in the lines and the oil, and you charge over the top of it.

Temperature works against you when it is cold. The colder the system and the surrounding air, the lower the water's vapor pressure, so you have to pull a deeper vacuum to make it boil, the slower the moisture comes off, and a cold-weather evacuation takes longer and can stall before it is dry. Warming the system, or at least the area around it, gives the moisture the energy to boil and move. Gentle heat on the lines, a warmer space, or simply more patience pulls water that a cold system holds onto. This is the physics the decay test catches: a system that levels off wet at low temperature has moisture the vacuum has not finished boiling away yet.

When do you triple-evacuate with nitrogen?

Triple evacuation is for a system you know is wet or contaminated, where one pull will not get the moisture out. You pull the system down to a vacuum, break the vacuum with dry nitrogen back up to a low positive or near-atmospheric pressure, then pull it back down, and you do that three times before the final deep pull and decay test. The nitrogen is the reason it works.

Here is the mechanism. A single deep vacuum boils off moisture, but in a wet system, in oil-soaked low spots and dead legs, the last of the water clings on and a single pull leaves some behind. Breaking the vacuum with dry nitrogen floods the system with a dry gas that mixes with the remaining moisture and dilutes it, and the next pull-down carries that diluted mix out. Each cycle takes out a chunk of what is left, and three cycles get a stubbornly wet system far drier than one long pull ever will. Use dry nitrogen, not air and not the refrigerant, because air puts moisture and oxygen back in and refrigerant is wasteful and illegal to vent later.

You do not triple-evac every job. A clean, dry new install with good brazing under nitrogen often pulls down and holds on the first try, and a single deep evacuation with a passing decay test is enough. Reach for the triple evac when the system was open to weather, when it burned out a compressor and you are flushing acid and moisture, when the decay test keeps leveling off wet, or when you have any reason to think water got in. The leak detection and recovery guide covers the nitrogen sweep and the burnout cleanup alongside it.

Pressure test with nitrogen before you pull a vacuum

Find the leaks before you evacuate, not after. The order that saves you time is to pressure test the system with dry nitrogen and prove it tight first, then pull the vacuum. Pulling a deep vacuum on a system that has a leak in it is a waste of pump time, because the vacuum will never hold and the decay test will fail, and you will not know whether the failure is the leak or moisture until you go back and pressure test anyway. Do it first.

Pressurize with dry nitrogen to the test pressure the equipment manufacturer specifies, isolate the nitrogen, and watch the pressure hold over time, correcting for temperature swings, because nitrogen pressure rises and falls with the air temperature on its own. A system that holds nitrogen is a system worth evacuating. A system that walks the pressure down has a leak you need to find and fix before the vacuum step. The leak detection and recovery guide covers the standing pressure test, the temperature correction, and where leaks hide in detail.

Nitrogen earns its place all through this work. You pressure test with it, you sweep with it on a triple evac, and you flow it through the tubing while you braze so the inside of the copper stays clean. Dry nitrogen is inert, dry, and legal to release, which is exactly why it is the gas for every step around the vacuum except the vacuum itself.

The filter drier as the backup, not the fix

Replace the filter drier whenever you open a system, because the drier catches the residual moisture the vacuum could not get and traps the small debris that comes with any repair. The drier holds a desiccant that adsorbs the trace water left in the system after a good evacuation, and a fresh one is cheap insurance that the last bit of moisture ends up locked in the drier instead of circulating to the metering device and the compressor.

The drier is the backup, not the substitute. A filter drier is sized to mop up the trace moisture left after a proper deep vacuum, not to dry out a system that was charged wet. Skip the evacuation and lean on the drier and you overwhelm it fast, the desiccant saturates, and the moisture moves on through. Worse, a drier that fills with water can break down and shed its desiccant into the system. Evacuate the system dry, then let the drier catch the remainder.

On a compressor burnout, where acid and moisture are through the whole system, a suction-line drier goes in alongside the standard liquid-line drier to clean up the circulating refrigerant, and it often gets changed again after the system runs a while. Match the drier to the refrigerant and the system, install it in the right direction, and change it any time the system has been open.

Why won't my vacuum pull down?

When a vacuum will not reach the target or takes forever to get there, the cause is almost always on a short list, and they rank by how often they bite. Work the list in order and you find it fast.

Restriction is first: 1/4 inch hoses and Schrader cores left in the ports strangle the flow, so the pump pulls down slow no matter how good it is. Contaminated pump oil is second, and it is the sneaky one, because milky or moisture-laden oil will not let even a healthy pump reach a deep vacuum. A leak is third, and a leak will stall the vacuum at a level it cannot get below, because air is coming in as fast as the pump pulls it out. A lot of moisture is fourth, because the boiling water gives off vapor that the pump has to keep clearing, which holds the reading up until the water is gone. And a big system or a long line set simply has more volume to evacuate, so it takes longer even when everything is right.

Before you blame the pump, cap its ports and check that it pulls itself down to its rated blank-off vacuum on fresh oil. If the pump is good and the setup is right and the vacuum still stalls high, you have a leak or a wet system, and the decay test reading tells you which. A stall that holds at a level is usually moisture. A reading that will not stop climbing is usually a leak.

System size, wetness, and the time it really takes

A bigger system and a wetter system both take longer to evacuate, and the most expensive mistake in this whole job is rushing it. Evacuation time scales with the volume you are pulling down and the moisture you are boiling out, so a small dry mini-split and a large commercial system with hundreds of feet of pipe are not the same call, and pretending they are is how a wet system gets charged.

The honest version is that the system tells you when it is done, not the clock. A small, dry, well-rigged system can reach a deep vacuum and pass a decay test in well under an hour. A large system, a long line set, or a system that has been open to humid weather can run the pump for hours and still need a triple evac to get dry. The decay test is the referee. If it levels off wet, the system is not done, regardless of how long the pump has already run.

Build the time into the job instead of fighting it. Bigger pump CFM speeds the pull-down on a large system, the right hose and core setup speeds every system, and warming a cold system speeds the moisture out. What you do not do is watch the pump for fifteen minutes, see a low number with the pump still running, and charge. The system that gets rushed is the system that comes back.

After the vacuum: breaking it and charging

Once the system holds a deep vacuum and passes the decay test, you break the vacuum with refrigerant, not with air, and move to charging. Reinstall the Schrader cores with the core tool without losing the vacuum, then introduce refrigerant to bring the system up off the vacuum. The first refrigerant in breaks the vacuum and from there you charge by weight or by the running readings.

The charging itself is its own job, and the charging guide covers it in full: weigh in the factory charge plus the line-set adjustment on a new or empty system, then verify by superheat on a fixed-orifice system or subcooling on a TXV or EEV system against the equipment targets. The short version is that the vacuum gets the system clean and dry and ready, and the charge gets it running right. Do not blur the two. A perfect charge on a system that was never properly evacuated is still a system that will fail.

The clean handoff is the point. Evacuation ends with a system that is dry, tight, and empty, proven by the decay test. Charging begins there. The record from this step, the vacuum reached and the decay result, belongs in the same file as the charge, because together they prove the system was both dry and correctly filled.

Recover first: EPA 608 and the no-venting rule

If there is refrigerant in the system before you open it, you recover it, you do not vent it. EPA Section 608 of the Clean Air Act prohibits knowingly venting refrigerant during service, maintenance, or disposal, and the prohibition covers the HFCs and the A2L blends, not just the old ozone-depleting refrigerants. Recover the charge into an approved cylinder with certified recovery equipment before any joint comes apart, and only then start the evacuation work this guide covers.

Recovery and evacuation are different operations that get confused because both move refrigerant or air out of the system. Recovery pulls the existing refrigerant charge out into a cylinder so it can be reused or reclaimed, and it happens before you open the system. Evacuation pulls a deep vacuum on the empty, open system to remove the air and moisture before charging. You recover, you repair, you evacuate, you charge, in that order. The leak detection and recovery guide covers the recovery step, the cylinder fill limits, and the 608 recordkeeping in detail.

Technicians who handle refrigerant must hold EPA Section 608 certification, and the recovery records are something an auditor can check after the fact. Recover it, log it, then evacuate the clean system. Venting to skip the recovery step is illegal, and on the A2L refrigerants it is a fire hazard on top of being against the law.

Large-system and data center evacuation

On a large system, the evacuation gets harder in every way that matters, because the volume is bigger, the pipe runs are longer, and there are more dead legs and low spots for moisture to hide. The setup that is merely helpful on a residential unit becomes the difference between finishing and not finishing on a large one. Pull from multiple ports, use the largest hoses you can, remove every core in the path, and bring enough pump CFM that the pull-down does not run into days.

On a data center, a hospital, or any critical-cooling load, the evacuation is also scheduled around an outage, because the cooling cannot be down while you work. The redundant capacity carries the room while you isolate, repair, and evacuate the failed unit, which is the reason those systems are built N+1 or better. The vacuum still gets pulled deep and the decay test still gets run, because a critical system charged over moisture fails the same way a residential one does, except the failure takes down a room full of equipment instead of a comfort load.

Time and patience are the trade on these jobs. A large or long-line system can need a triple evacuation and hours on the pump, and the pressure to get the room back online is exactly the pressure that tempts a crew to charge a wet system. The commissioning record on these systems is part of the asset, so the vacuum reached and the decay result get documented like everything else, because the next tech inherits the consequences of a step that got skipped.

What to document

A vacuum is invisible the moment refrigerant goes in over it, so the only proof the system was pulled deep and held is what you write down now. Six months out, when a system runs low on capacity or a compressor fails with acid in the oil, the record is what answers whether the system was ever properly dried. The vacuum reached and the decay test result are the proof, and without them the next tech is guessing about a step they cannot see.

Capture the deepest vacuum reached, where the micron gauge was placed, the decay test result over the hold time, whether you ran a single or a triple evacuation, the pump and the condition of its oil, whether you changed the filter drier, the nitrogen pressure test result that preceded it, and who did the work and when. The table below is the short version. Tie it to the charging record so the file shows the system was both dry and correctly charged.

Common mistakes

• Guessing the vacuum off the manifold compound gauge instead of reading it on a micron gauge.
• Leaving the Schrader cores in and pulling through 1/4 inch hoses, so the vacuum is slow and never gets deep.
• Reading the micron gauge at the pump instead of on the system, so the number looks better than the system ever reaches.
• Calling it evacuated when the pump touched 500 microns, without isolating the pump and running a decay test.
• Running contaminated or low pump oil that will not let the pump pull down, and blaming the pump or the system.
• Pulling a vacuum on a system that has a leak, instead of nitrogen pressure testing and fixing it first.
• Charging a wet system because the evacuation got rushed to get the unit back online.
• Leaning on the filter drier to dry a system that was never properly evacuated.

Field checklist

Standards and references

The equipment manufacturer governs the evacuation. The install and service manual sets the vacuum target for the specific unit, and while 500 microns or below is the common figure across the trade, the number that controls is the one the manufacturer specifies, so confirm it for the equipment in front of you. Some systems and oils call for a deeper vacuum. The manufacturer also sets the nitrogen test pressure, the factory charge, and the filter drier specification.

EPA Section 608 of the Clean Air Act, at 40 CFR Part 82, Subpart F, controls the refrigerant side of the work: it prohibits knowingly venting, requires recovery before a charged system is opened, and requires technician certification. It covers the HFCs and the A2L blends, not just the older refrigerants. The recovery happens before the evacuation, so the 608 obligations attach to the front of this job.

Industry guidance on deep evacuation and the decay test comes from ACCA and ASHRAE on the install and commissioning side, and from the micron-gauge and vacuum-pump manufacturers on the tools and the procedure. The decay-test criteria, the acceptable rate of rise, and the recommended hold time vary with the source and the system, so follow the equipment manufacturer's evacuation procedure and the gauge maker's decay-test guidance, and confirm any code requirements the jurisdiction has adopted. Cite the body that owns the point, and treat the 500-micron figure as a common target hedged to the manufacturer's spec, not a universal mandate.

Units, terms, and conversions

Vacuum and the related readings show up in a few unit systems, so the same idea reads differently across a manifold, a micron gauge, and a manual.

Deep vacuum is read in microns, where 1 micron is one-thousandth of a millimeter of mercury and atmospheric pressure is about 760,000 microns, so 500 microns is a very deep vacuum. The same vacuum can be written in microns of mercury or, more coarsely, in inches of mercury (in Hg) on the manifold's compound gauge, which only resolves the shallow end. Nitrogen test pressure reads in psig, the gauge pressure above atmosphere. Pump displacement is in cubic feet per minute (CFM), and nitrogen purge flow for brazing reads in cubic feet per hour (CFH). Refrigerant charge is tracked in pounds and ounces in the field and kilograms in metric data.

Evacuation

Pulling a deep vacuum on an open system to remove air, moisture, and non-condensables before charging

Micron

Unit of deep vacuum, one-thousandth of a millimeter of mercury; 500 microns or below is the common target

Micron gauge

Electronic vacuum gauge that reads the deep-vacuum range; placed on the system, away from the pump

Decay test

Isolating the pump and watching the micron gauge for a rise that signals a leak or remaining moisture

Triple evacuation

Pulling vacuum, breaking it with dry nitrogen, and repeating three times to dry a wet or contaminated system

Non-condensables

Gases such as air that will not condense in the system and raise head pressure until evacuated out

Filter drier

A desiccant-filled component that catches the trace moisture left after a proper vacuum; a backup, not a substitute

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• Do I charge by superheat or subcooling?
• How do you find a refrigerant leak?
• What tools does an HVAC technician need?
• Why do you flow nitrogen when brazing refrigerant lines?
• What is a metering device in HVAC?