Aircon refrigerant charge: the correct amount is a weight

A refrigerant charge is a quantity, stated as a weight. The manufacturer fixes it for each model and prints it on the outdoor unit, in grams or in kilograms depending on the size of the system. It is not a level to be judged by eye. It is not a setting anyone dials in on site.

The figure attaches to a model, and to nothing broader than that. A brand name settles it no more than a category does. Coil volume, pipe bore and compressor size all shift across units that appear identical in a showroom. A machine built around one weight will not behave correctly holding a neighbouring model's weight, which is why no honest source publishes a single number for aircon in general.

The nameplate settles it. That plate sits on the side or rear panel of the outdoor unit, beside the model number and the refrigerant type. The installation manual repeats the same figure and adds the conditions attached to it. Both are ordinary published documents, so the number exists on paper long before anyone connects a hose to the system.

What the plate states assumes a reference length of pipe. The factory seals the charge in before the unit ships, sized for the machine itself plus a stated run of pipework. Where the actual route is longer than that reference, the correct quantity for the installation is larger than the plate says. The printed figure starts the calculation rather than finishing it.

Pressure inside a running system is not a count of what is in it. Pressure reports the condition the refrigerant is in at that moment, and the condition moves with the weather, the room, and the effort the compressor is putting in. One correct charge reads high at noon on a west-facing ledge and lower at night in a shaded corridor. Nothing entered or left the circuit in between.

Heat load moves the reading just as much. A flat switched on after a hot afternoon presents the system with a large load, and pressures sit high while that load comes down. The same machine holding a settled room reads lower on the same gauges. Both readings are correct. Neither is a quantity.

Weight holds still. Refrigerant does not thin out, does not evaporate away inside a sealed circuit, and does not weigh differently in the afternoon. A stated mass is a fixed target, and a fixed target is something a homeowner can check against, argue with, and find written on an invoice afterwards.

None of this is an argument against gauges. Pressure readings taken alongside pipe temperatures produce superheat and subcooling, and that pair is how a technician judges whether a running system is being fed properly. What gauges cannot do is count. They describe how the circuit is behaving. Only a scale establishes how much is inside it.

Pipework holds refrigerant, so a longer route needs more of it. The manual states a rate for exactly this, given as a weight per metre of pipe past the pre-charged length. The sum is arithmetic. Routed length, less the pre-charged length, multiplied by the stated rate, produces a figure in grams that gets weighed in on a scale.

Seeing extra refrigerant itemised against a long route is a good sign rather than a suspicious one. That entry is not padding. The unit was designed around a reference length the flat did not provide, and the shortfall has to go in for the machine to deliver what it was rated at. Alongside the price, the entry should name the weight and how it was arrived at.

Leaving it out puts the system short on the day it is commissioned. Handover exposes none of that. Cold air comes out of the grille, the flat cools down eventually, and a brand-new installation is given the benefit of every doubt going. Months later the weak room gets blamed on afternoon sun, on the ceiling height, or on the brand printed on the box.

Whether the route is allowed at all is a separate question with its own answer. Pipe run length limits govern how far the outdoor unit may sit from the room it serves. The additional charge governs whether a permitted route was commissioned correctly. A run can clear its limit comfortably and still be under-filled, because the two get checked at different moments by different people.

Both conditions cool badly and they arrive there by opposite routes. Too little refrigerant starves the indoor coil. Too much floods it. Which one is in front of you decides whether gas comes out or goes in, and a wrong guess costs something in either direction.

The starved version is the familiar one. Pressure in the indoor coil falls, and the coil surface can drop below freezing and ice over. Airflow across an iced coil collapses, so the room gets worse rather than slowly better. Circulation also carries heat away from the compressor, and a thin circuit leaves that machine working at a temperature it was never meant to hold.

The flooded version is the one nobody expects, because more sounds as though it should mean colder. It does not. Surplus refrigerant fills part of the outdoor coil that should be rejecting heat, so less heat leaves the system and pressures climb. The compressor pushes harder against that resistance for less cooling in the flat, and the electricity bill rises while comfort falls.

The heaviest consequence lands on the compressor itself. Liquid that should have boiled off indoors can reach the compressor inlet instead, and that machine was built to squeeze vapour. Liquid does not compress. Repeated slugs of it batter valves and bearings, and the damage builds quietly behind a system that is still blowing cold air.

Neither condition separates itself at the vent. A flat that cools slowly feels identical in both cases, which is why the direction gets decided at the scale and the gauges rather than in the room.

Adding gas until the air feels cold aims at a sensation rather than a quantity. The sensation arrives before the correct weight does, and it carries on arriving after the correct weight has passed. At no point in that process does the system announce that it now holds enough.

Overshoot is the natural result. Cooling improves steadily while a short circuit fills, so the feedback keeps saying yes. Past the specified weight the feedback goes quiet instead of reversing, because a mildly over-filled system still blows cold air on the day. Whatever went in past the line stays in.

The method also builds on an unknown base. Nobody in the flat knows what the circuit held before the hose was connected, so nobody knows the total afterwards either. The next technician inherits a system with no established quantity in it, and the same guess starts over from a different unknown.

Why the circuit was low in the first place is the question that decides whether any of this repeats. Refrigerant is not used up by running, so a system short of it has lost it somewhere. Restoring the weight without locating the opening resets the clock and settles nothing.

Putting the charge back after a confirmed repair is a different act entirely, and it is the normal end of that job. The circuit gets emptied, the fault gets fixed, and the specified weight goes back in. That sequence contains a quantity from start to finish. A blind top-up contains none.

A correct recharge begins by emptying the circuit rather than adding to it. Whatever is in there is an unknown amount, and an unknown plus a known is still an unknown. Recovery draws the existing refrigerant out into a cylinder, which is also how it is kept out of the atmosphere rather than vented into it.

Pulling a vacuum comes next. Air and moisture get into any circuit that has been opened, and both have to come out before refrigerant goes in. The vacuum and pressure test is a subject in its own right, with its own readings and its own hold times. What the quantity depends on is the order. An empty, dry, proven circuit is what makes a measured amount mean anything.

The charge goes in by weight after that. The cylinder sits on a scale, the target is the nameplate figure plus whatever the pipe route adds to it, and the scale decides when to stop. The number is settled before the hose is connected. Charging a system and topping one up part company right there.

Verification comes last, and it never replaces the scale. Superheat and subcooling, read against the conditions on the day, confirm the machine is behaving as it should with that weight inside it. A reading that disagrees with a correctly weighed charge is pointing at something else in the circuit. That is useful information, not an invitation to add more gas.