What it is

Evacuation is how you get the air and moisture out of a system before you let refrigerant in. You pull a deep vacuum with a vacuum pump, watch it on a micron gauge, hit your target, then close the system off the pump and run a decay test to prove it's both dry and leak-free. Air and water are the two enemies of a refrigeration system, and a proper evacuation is the only way to remove them.

How it works

Two problems live inside a freshly brazed system: non-condensable gases (mostly air, which won't condense and just raises head pressure and superheat error) and moisture (water vapor, which reacts with POE oil to form acid and can freeze at the metering device). You can't sweep them out with refrigerant — you have to pull them out with vacuum.

Here's the part techs miss: water boils at low temperature when the pressure is low enough. At a deep vacuum, any liquid water in the system flashes to vapor and the pump can carry it away. But that takes time and a genuinely deep vacuum. A shallow vacuum leaves the water as a liquid, hiding, ready to wreck the compressor later. That's why the target is measured in microns, not in inches of mercury — and why a standard compound gauge is useless here. The gauge needle doesn't even leave the peg in the range that matters.

In the field

  1. Connect for flow, not for choke. Use large-diameter hoses or a vacuum-rated hose set, and pull from both the high and low sides at once. Remove the Schrader cores with core-removal tools — the tiny core orifice can quadruple your evacuation time all by itself.
  1. Put the micron gauge at the system, not at the pump. The deepest, cleanest vacuum is right next to the pump. The system itself is always a little higher. Read the micron gauge as far from the pump and as close to the system as you can — that's the number that matters.
  1. Pull down to target. Run the pump until the micron gauge settles at or below 500 microns. If it stalls high and won't come down, you have a leak, trapped moisture still boiling off, or a restriction in your hoses/cores.
  1. Valve off and run the decay test. Close the valve that isolates the system from the pump (or close the core tools) and watch the micron gauge with the pump removed from the equation. Don't just shut the pump off — isolate the system.
  1. Read the rise. A good system holds. A small, slow rise that levels off means residual moisture still outgassing — keep pulling. A steady rise that doesn't level off means a leak. Big difference, and the decay curve tells you which.

Normal values & targets

  • Evacuation target: pull to 500 microns or lower for a standard residential install. Many techs shoot for 250–300 to give themselves margin.
  • Decay test pass (typical): after isolating, the vacuum should hold below roughly 500 microns and rise no more than a small amount — a common field standard is staying under about 500 microns, or rising less than ~250–300 microns, over a several-minute hold. Confirm any spec the manufacturer gives.
  • Atmospheric pressure for reference: about 760,000 microns at sea level — so 500 microns is a deep, deep vacuum.
  • Blank-off test of the pump alone: a healthy pump with good oil should pull well below 100 microns at the pump with the gauge right on it.

Common faults & what they mean

  • Vacuum stalls around 1,500–5,000 microns and won't drop: usually moisture still boiling off, a small leak, or restricted flow (cores still in, skinny hoses). Let it run and watch — moisture eventually breaks free; a leak never improves.
  • Pulls down fast, then rises steadily and never levels off on the decay test: leak. Find it and fix it before you charge.
  • Pulls down, rises a little, then flatlines: moisture outgassing. Keep evacuating; it'll get there.
  • Can't get below a few thousand microns ever: check your pump oil first — contaminated oil kills ultimate vacuum. Change it and retry.

Tech tips & gotchas

Old vacuum pump oil is the number-one reason a pump won't pull deep. Oil absorbs the moisture you're removing and goes bad fast. Change it often — sometimes more than once on a wet system — and always change it after evacuating a burnout.

A standing decay test is non-negotiable on a quality install. Hitting 500 once and immediately charging tells you the pump reached 500, not that the system will hold it. The decay test is the proof.

Schrader cores are vacuum killers. Pulling through a core can turn a 20-minute evacuation into a two-hour one and can mask a slow decay. Use core removal tools so you're evacuating through a full-bore port.

If you have a real leak you can't find, pressurize with nitrogen and a trace of refrigerant and leak-detect — don't try to chase it in a vacuum. You evacuate to remove moisture and air; you pressure-test to find leaks. Two different jobs.

Cold weather hides moisture. If it's freezing out, the water in the system may be ice and won't boil off until things warm up — your vacuum will look great and then the moisture shows up later. Gentle warmth on the suspect area helps drive it out.

Safety / code notes

  • A vacuum pump exhausts whatever it pulls — including any residual refrigerant. Recover refrigerant per EPA 608 before you ever pull a vacuum on a system that had charge in it; you don't evacuate refrigerant to atmosphere.
  • Never use the system's own compressor to "pull a vacuum." Running a compressor in a vacuum can arc the internal motor terminals and destroy it.
  • Wear eye protection when breaking vacuum and admitting refrigerant — and admit charge slowly.
  • Proper evacuation supports the equipment listing and good-practice mechanical code (Indiana enforces the IMC); a system charged wet or with non-condensables will not perform to its rating and can fail prematurely.