Overcharge Is a Fault Too, and About a Quarter of Systems Have It

Refrigerant gets discussed as though the only failure mode is running out. The field data does not agree. NREL's pooled analysis of more than 2,600 residential systems across eleven field studies found roughly a quarter undercharged and roughly a quarter overcharged.

How it happens

Almost always the same way: somebody added refrigerant on a pressure reading rather than weighing the charge in. Pressure is easy to read and it responds immediately, which makes it a tempting proxy. It is a poor one — pressure moves with ambient, with load, and with airflow, so a system that reads low on a mild morning may be perfectly charged.

The result is a system with more refrigerant than the manufacturer specified, and nobody ever comes back to check, because the complaint that prompted the visit went away.

What it costs to run

NIST's instrumented measurements put a 30% overcharge at about 6% more total power for roughly 3% lower COP. That is real money paid every cooling hour, but it is a milder penalty than the undercharge case — the same test measured 30% undercharge at 12% lower COP and 15% lower capacity.

So on the electricity bill alone, an overcharge is the less urgent of the two. That is not the whole story.

What it does to the compressor

The mechanism worth understanding is liquid. Excess refrigerant has to go somewhere, and it stacks up as liquid in the condenser, reducing the surface actually available to reject heat. Head pressure rises. On the hottest days, when the condenser is already struggling, it has least margin.

The bigger concern is off-cycle. Surplus refrigerant migrates to the coldest part of the system while the machine is stopped, and a compressor that starts with liquid in it is doing something it was not built to do. Liquid does not compress. This is well described qualitatively in the literature and, like almost everything about equipment life, nobody has published a number for it.

Why it is hard to catch

An overcharged system cools. Often it cools well. There is no comfort complaint, no iced coil, no obvious symptom — which is exactly why a quarter of systems can carry it indefinitely.

There is a subtler trap for anyone monitoring: a system that was already overcharged when monitoring began has no deviation from its own normal to detect. It learns the fault as normal and becomes permanently invisible. Catching that case needs an absolute test rather than a relative one — the signature of a growing liquid column, with the condenser split staying flat as load rises.

The honest advice

If a technician adds refrigerant to your system, it is entirely reasonable to ask two things: how much went in by weight, and where it went. Refrigerant is not consumed. If a correctly charged system needed more, something let the last lot out.

And if the correction is a mild overcharge with no other symptoms, the sensible answer is usually to bundle it into the next scheduled visit rather than pay for a truck roll — the running cost rarely justifies the call on its own. The reason to fix it is the compressor, not the bill.