Aircon pump down explained: saving the existing charge
A relocation quote rarely names it, but the pump down decides whether the refrigerant already in your system survives the move. Done poorly, the charge escapes without a sound, and the recharge that follows bills you for gas that was already yours.
By Team Snowflake | Updated 5 Aug 2026
What a pump down moves, and where it goes
A pump down clears the refrigerant out of the pipework before anyone opens it. The charge is driven into the outdoor unit and shut in behind the two service valves on its side. The connecting lines and the indoor coil are then empty enough to disconnect without losing the gas.
Nothing is vented and nothing is bought. The refrigerant circulating that morning is the same refrigerant that goes back round the loop once the pipes are rejoined. A system that holds on to its own charge does not need a recharge afterwards.
This sits underneath almost every relocation and most indoor unit replacements. Moving a wall unit to a different wall, shifting the condenser onto a new bracket, or swapping an indoor unit on an otherwise healthy system all require the pipework to come apart. Each one begins by settling what happens to the gas.
Owners meet the term without ever meeting the explanation. It turns up as a line on a quote, or as a word said on the phone while a job is being scoped, and nothing in it tells a homeowner what is being promised. Reading that line properly starts with knowing what the step is for.
Pump down and evacuation are opposite operations
These two are confused more than any other pair in the trade, and they do opposite jobs. A pump down protects refrigerant that is already in the system, gathering it somewhere safe so the work can happen around it. Evacuation strips a circuit that has stood open, drawing out air and water vapour so a charge has clean space to enter.
The order tells them apart. A pump down comes first, while the system is still whole and still holds its gas. Pulling a vacuum comes last, after the pipes are back together and before anything goes in. A relocation done properly involves both, and asking about only one of them leaves half the job unexamined.
Why the charge is saved instead of vented
Refrigerant costs real money, and the quantity sitting in a domestic split is not a rounding error. Letting it go at the start of a job means buying it back at the end of one. That cost lands on the customer, which is what turns a technical step into a commercial one.
Release is also the wrong thing to do with it. R32 and R410A both carry a warming effect well above carbon dioxide, and a charge vented off a ledge does not disappear. Handling refrigerant properly belongs to the trade rather than to courtesy, and a technician who treats venting as routine is telling you something about the rest of the work.
The third reason is the least obvious. A system that keeps its own charge never has to have that charge decided again. Weighing a fresh charge in correctly takes care and a set of scales, and every recharge is another chance to end up slightly over or slightly under. Holding the original quantity avoids the question altogether.
Recovery is the other way to empty a circuit
A pump down is not the only honest method, and the alternative is worth knowing. Recovery draws the refrigerant out into a cylinder using a dedicated machine, rather than pushing it into the condenser. The pipework ends up clear either way. What changes is where the gas sits while the work happens.
Recovery is the right call when the outdoor unit cannot be trusted to hold the charge. Worn valve seats, a compressor that will not run, and a system already losing gas all rule out a pump down. What matters to an owner is that one of the two gets named before the job starts, because the third option is simply letting it go.
How the refrigerant gets shifted, in principle
The method rests on one fact about the circuit: refrigerant travels one way round it. With the system running in cooling, the compressor pushes liquid out through the thin line towards the indoor coil, and draws vapour back through the larger insulated one. Shut the outward path and the return path is still open.
So the liquid valve closes first, with the system still running. The compressor keeps drawing refrigerant out of the indoor coil and the connecting lines, and nothing is going out to replace it. The charge gathers in the condenser and the outdoor pipework, which has room to hold it.
What gets watched during that is pressure, not the clock. The gauge on the suction side falls as the lines empty, and the suction valve is closed at the point where the reading says the refrigerant has arrived. That single judgement is where the procedure is won or lost. The rest is fittings and spanners.
None of this is a homeowner procedure, and the account above is deliberately not a method. It reaches inside a sealed circuit under pressure, it depends on gauges to know when to stop, and every way it goes wrong is silent at the time. Understanding it is worth something because it lets you ask about it.
Where a pump down goes wrong
Most losses come from a short list, and none of them look like anything on the day. A bad pump down and a good one leave the same picture behind. An outdoor unit with its valves shut, and open pipework beside it. The difference is on the inside of the valves.
Closing the valves in the wrong order is the plainest failure. Shutting the suction valve first traps refrigerant in the indoor coil and the connecting lines instead of clearing them, and that trapped charge leaves through the first flare that gets broken. It goes fast and it goes quietly.
Running past the point of collection costs the compressor rather than the gas. Once the low side is pulled into vacuum, the compressor is working against nothing, and pressure inside the circuit sits below the pressure outside it. Air is then drawn in through any joint or seal that is less than perfect, and moisture arrives with it.
Valve seats that no longer seal are the failure with a delay built into it. The valve reads as closed, the spanner says closed, and the seat is passing refrigerant slowly back into the open pipework. Older units, valves that have been worked many times, and stems over-tightened at some point are the usual candidates. Left overnight between disconnection and reconnection, a slow bleed empties a system completely.
| Where it fails | What becomes of the charge | When it surfaces |
|---|---|---|
| Where it failsThe valves are shut in the wrong order | What becomes of the chargeRefrigerant stays behind in the indoor coil and the lines | When it surfacesAt the moment the first flare is broken, and there is no getting it back |
| Where it failsThe system runs on past the point of collection | What becomes of the chargeThe low side goes into vacuum and pulls outside air in through weak joints | When it surfacesMuch later, once moisture in the circuit has worked on the oil and the windings |
| Where it failsA valve seat is worn and no longer seals | What becomes of the chargeThe charge bleeds slowly back out into the open pipework | When it surfacesOn reassembly, after the system has stood disconnected |
| Where it failsThe system was already leaking before anyone arrived | What becomes of the chargeIt collects as it should, then escapes through the opening it always had | When it surfacesAt restart, as cooling that never fully returns |
| Where it failsA pipe is cut before the gauge confirms the low side is clear | What becomes of the chargeWhatever was still in that line goes straight to the air | When it surfacesImmediately, though only the person holding the cutter can tell |
A leaking system cannot be pumped down
A system already losing refrigerant will not hold what gets pushed into it. The charge gathers in the condenser as it should, then bleeds out through the same opening it was escaping from all along. Nothing about the procedure fails visibly. There is simply less in there than anyone assumed once the pipes go back on.
So a leak found before a move changes the job rather than delaying it. The sensible order is to establish whether the system holds pressure at all, then decide whether the charge is worth saving. A relocation booked on a system that has been topped up more than once should start with that question.
Why the result stays hidden until restart
There is nothing for an owner to see. A closed valve cannot be read from outside, and a disconnected indoor unit says nothing about how much refrigerant is sitting in the condenser out on the ledge. The charge is either in there or it is not, and the answer arrives when the system goes back together and gets switched on.
That blindness is what lets the cost move quietly onto the bill. A relocation finishes, the unit blows cold on the day, and nobody counts what went into it. Where the charge was lost, the shortfall surfaces later as cooling that fades in the heat of the afternoon, or a room that never quite reaches the setting.
A gas top-up appearing at the end of a relocation deserves a question before it deserves a payment. The refrigerant in that system belonged to the owner at the start of the day. If it is being sold back, something happened to it in between, and the pump down is where that would have happened.
The honest reasons gas gets added
Some relocations genuinely need refrigerant, and telling those apart from the rest is most of the skill in reading a bill. A longer pipe run holds more refrigerant than a short one, so moving a condenser further from the indoor unit raises the charge the system needs. That is arithmetic rather than an upsell.
A system that was low before anyone touched it is the other real case. So is one where recovery was used and the recovered gas is not clean enough to go back in. Each of those carries a reason that can be stated in advance. The test is not whether gas is charged for. It is whether the reason was named before the work rather than after it.
What to ask when a relocation is quoted
Ask before the job, and ask about method rather than intent. Nobody responds well to being accused of planning to vent your gas, and an installer who does this properly is usually happy to walk through the method. The questions worth asking are the ones that can only be answered with a step, a reading, or an order of operations.
The answers sort themselves quickly. A method-shaped reply names what happens to the refrigerant, what tells the technician it has moved, and what the fallback is if the valves turn out not to hold. A reply built on a general commitment to careful work has addressed none of that.
None of this asks an owner to understand gauges. It asks the installer to describe something that was going to happen anyway. Where the description does not exist, the plan usually does not either, and a relocation quote that has not priced the refrigerant is a quote with an open end in it.
| Question worth asking | A reply with method in it | What a thin reply leaves open |
|---|---|---|
| Question worth askingHow will the refrigerant be saved before the pipes come apart? | A reply with method in itA pump down into the outdoor unit, or recovery into a cylinder, named as a step of its own | What a thin reply leaves openIf neither gets named, the working plan may be to break the flare and let the gas go |
| Question worth askingWhat tells you the charge has collected? | A reply with method in itA gauge on the suction side, watched until the reading says the lines are clear | What a thin reply leaves openA reply built around how long the unit runs describes a habit, not a measurement |
| Question worth askingWhat happens if the valves turn out not to hold? | A reply with method in itThe system gets emptied into a cylinder instead, and you hear about it before that changes | What a thin reply leaves openWorn seats are common on older units, so a plan with no fallback will be improvised on the ledge |
| Question worth askingWill the pipework be vacuumed before the gas goes back? | A reply with method in itYes, with a reading, because the lines were open to the air for the whole move | What a thin reply leaves openCharging straight onto an opened circuit carries the room's moisture in to the compressor |
| Question worth askingIf a top-up is needed at the end, when will I know? | A reply with method in itBefore the work, with the reason named: a longer run, or a charge that was already low | What a thin reply leaves openA gas charge that shows up only on the final bill has no agreed reason behind it |
Ready to get started?
Tell us what’s going on. Symptoms, setup, photos, anything we should know. We’ll assess and come back with the right next step.