Join 3,562 readers in helping fund MetaFilter (Hide)


Why does a low refrigerant charge cause an evaporator to ice up?
June 22, 2013 4:44 PM   Subscribe

I have read that a low refrigerant charge in a refrigeration system can cause the system's evaporator to ice up. I happen to have a dehumidifier which is exhibiting this symptom. I do not expect to be able to repair the dehumidifier but, as something of an HVAC enthusiast, I am curious as to why and how this happens.

I am not by any means an HVAC technician, although I would like to understand refrigeration systems better than I do now.

I have a dehumidifier here which is definitely not suffering from a dirty evaporator coil or restricted airflow over the evaporator coil. It is currently operating in a room in which the temperature is 76 degrees Fahrenheit and the relative humidity is 58%. Nevertheless, the bottom part of the evaporator ices up after only a few minutes of operation:

Picture of Iced Evaporator

Of all the explanations I have read for why a dehumidifier's evaporator might ice over, only the "it has a slow refrigerant leak and hence a low charge" explanation seems to apply. But I don't understand why a low refrigerant charge would result in a portion of the evaporator becoming excessively cold, and I hope someone might be able to explain it to me.

For what it's worth, this dehumidifier was purchased in or around 2004, and I believe (but am not 100% sure) that it uses R-134a refrigerant.

Again, I'm not looking to repair the unit. Rather, I only want to learn more about how these refrigeration systems work and what might be causing this to happen.
posted by Juffo-Wup to Science & Nature (6 answers total) 1 user marked this as a favorite
 
The trick here to note that when your refrigerant pressure is low (56-58 psi for R-134a), the vaporization temperature is 0C. Which is conveniently the freezing temperature of water; and you'll start freezing out water from the air. The vaporization temperature (or saturation temperature) and pressure are always related. If you were at a higher refrigerant pressure, the evaporator would be removing heat at a higher temperature, hence no ice will form.
posted by freshkippers at 6:46 PM on June 22, 2013 [1 favorite]


I had it explained to me that you're basically putting more cold into less refrigerant, dropping it below freezing when it should be above that at its lowest point.

And once the ice forms, it's a runaway process, because the ice insulates the coolant tubes when they are at their coldest, preventing them from shedding cold (or more accurately, absorbing heat) at that time.
posted by Hatashran at 9:15 PM on June 22, 2013


See the pressure/temperature chart for 134:

At 28 psig and the saturated temp is 32f and thus moisture from room air condenses on the coil and freezes instead of draining away. You don't have enough refrigerant in the low side of the cycle to absorb enough heat to maintain a condensing pressure high enough on the high side to feed the proper rate of liquid refrigerant through the metering device back into the low side... so the low side pressure drops meaning the temperature drops below freezing the the whole thing falls apart.

In air conditioning or dehumidification the coil must stay above the freezing point of water because there is no defrost cycle.

Now, as for why only part of the coil freezes - the for dummies version is liquid refrigerant flows through a metering device into the evaporator where the liquid boils while absorbing heat. A change of state (liquid to vapor) absorbs massive amounts of heat content, thus cooling the coil. Cooling the air passing through the coil lowers it's ability to hold moisture, thus condensing water drops out.

Your coil is frosting only for the first 1/3 because it is running out of liquid refrigerant. Normally the coil would boil off the last liquid to vapor in the last few tubing passes. You're boiling off the last bit of liquid 3 rows in. If you put a temperature probe on the unfrozen section you would find it is well above freezing, the refrigerant in those tubes is superheated vapor.

Source: I'm a Journeyman Refrigeration Mechanic

Thanks to the AskMe spell check I now know that I have been spelling evaporator wrong on my work orders for my entire career. Oh well, why stop now.
posted by vonliebig at 9:38 PM on June 22, 2013 [3 favorites]


My stab at simplifying it:

Less refrigerant means it can pick up less heat out of the air, so the little bit of refrigerant that remains gets super cold, picks up a little heat and then freezes.

In normal operation, the refrigerant stays above the freezing point because it is picking up all that moisture, which condenses and releases its heat to the refrigerant.

Basically, what's happening is exactly what happens with one of those spray air duster things. (*) When you start out, it blasts air out and the can starts getting really cold, really fast. But its ability to pull heat out of the air starts to diminish and it gets colder and colder, and the flow of air out of the nozzle slows down. Ice forms on the outside of the can

(*) In most cases, the "air" inside the cans is actually the same r134a or r152 that air conditioners use.
posted by gjc at 10:16 AM on June 23, 2013


Excellent; thanks vonliebig! If you're still reading, a couple supplemental questions:

1. In normal operation (i.e. with a proper charge), if there is liquid refrigerant in most of the evaporator coil, what prevents liquid refrigerant from entering the compressor and damaging it? (Or would it not cause damage so long as enough vapor is in the compressor too?)

2. If I were to continue running this dehumidifier (which I will not) and it continued leaking refrigerant, what is the end game? Is it just that eventually the pressure in the circuit would equalize with atmospheric pressure and there would be so little refrigerant left that almost nothing would get cold? Or would this condition cause the compressor to die (and if so, of what? Is there oil in the refrigerant? Or something else...)

Thanks again!
posted by Juffo-Wup at 3:22 PM on June 23, 2013


I just got home from a double shift so this may be a bit hard to follow, but stay with me:

1. You said it yourself, the key phrase here is "most of the evaporator coil." The plan in a Direct Expansion coil (which this is) is to use as much of the coil as possible for boiling refrigerant as possible.... without overshooting and sending liquid back to the compressor. This is because latent heat (heat absorbed in the change of state) absorbs way more energy that sensible heat (increase in temperature). Classic example is boiling water.... it's only one BTU to raise one pound of water one degree f... say from 200 to 201 degree f. Boiling one pound of water is about 970 BTU (212 degree f to 212 degree f). So once all the liquid in the coil is vaporized and we're just raising the temperature of vapor, we aren't doing all that much. We want to absorb lots of BTUs.

Liquid refrigerant in the compressor, or "slugging" as it's called, destroys the compressor. A refrigeration compressor is built to compress vapor, not liquid.... in the battle between metal and hydraulics, hydraulics wins every time. I've seen pistons the size of a coffee can blown to little pieces due to this, never mind the valves, backer plates, etc. Some designs, like Scroll compressors, can handle a little liquid without damage. The most common type, reciprocating (like what's in your home fridge) are a middle ground. Rotary (probably what is in your dehumidifier) can not handle any.

That's if you fill the compressor with enough to get liquid in the actual compressing part.... it grenades the internals. If you just get liquid in the compressor but not so much that you grenade it you'll seize up instead because liquid refrigerant is a solvent..... the oil will mix with it and be flushed out when the refrigerant boils in the crankcase.... lots of foam, not much lubrication.

So to finish answering the question, what keeps us from sending liquid back to the compressor is a well tuned system. Most of the coil is for latent heat.... it will be all the same temperature if you measure the tubing passes. That's where the refrigerant is a saturated mixture of liquid and vapor - it's boiling. The last few passes start raising in temperature. There is (hopefully) only vapor now, and it's picking up sensible heat, AKA superheat. 5-15 degrees is common, depending on application. That 5-15 degrees represents a compromise. We're wasting coil capacity to ensure the compressor does not shit the bed.

The metering device controls how much refrigerant ends up in the coil, and thus the superheat. For your dehumidifier it's probably a capiliary tube (look for a long skinny coil of copper tube going into the coil - that restricts the flow. Some engineer takes a guess at what temperature your room is at, figures out the airflow of the fan, and calculates how much refrigerant that load can boil off. Then he looks at the chart for his refrigerant and cuts the capiliary tube to length. Shorter is less restriction and more flow, longer is less flow.

If all else fails, some systems prone to liquid flood back to the compressor have accumulators (basically a buffer tank - liquid falls into it, but the compressor sucks off the top so unless you fill it completely you can only get vapor). Look and see if there is a little pill bottle sized can welded to the side of the compressor that the low side pipe feeds into, if it does your dehumdifier is a rotary compressor. That can is a tiny accumulator to try and protect it.

2. Assuming it's a piece of crap domestic appliance (most are) it keeps running even once it's out of gas. If the leak is on the low side it'll suck in some air too - that compressor will pull a 10-20 inches of mercury vacuum. Yes it will stop getting cold.

The compressor however will get very hot. The compressor is partially cooled by the refrigerant vapor flowing through it, so it will eventually overheat. Maybe the bearings seize and it flips the house circuit breaker. Maybe it rides the internal temperature overload on and off until the wires on the on off switch finally overheat and melt. Maybe it blows out the compression mechanism and just keeps running not doing anything.

More sophisticated systems have a low pressure safety switch. At a certain abnormally low pressure it just turns off until someone fixes it.
posted by vonliebig at 8:43 PM on June 27, 2013


« Older Hiring managers, have you ever...   |  That's about it - it was a one... Newer »
This thread is closed to new comments.