Join 3,416 readers in helping fund MetaFilter (Hide)

Broken Fridge maybe/maybe not?
September 8, 2011 7:09 AM   Subscribe

Broken fridge; need refrigeration engineer help for insurance and maintenance purposes. Many technical geeky issues involved. Long post follows:

My house came with an expensive Subzero 550 refrigerator/freezer.

During our recent Vermont floods, power was off for two days. The refrigeration compartment got to room temp (about 65F), with a medium payload. We were out of town during the outage.

The freezer compartment was set to –12F, and its contents stayed fairly frozen, if I use the ice tray as an indicator.

After the power returned, which was just before we did, the fridge failed to attain temperature, and when I measured it with an Omega HH303 thermometer, it showed about 60F. After a full night of operation, it was still in the high 40’s.

Calling a repairman, he blamed the problem on a clogged evaporator, which replacement I am currently awaiting. (BTW, Subzero/Wolff Company is providing this at no charge, out of warranty! Nice touch, folks. I am impressed!)

The tech showed me a pressure reading of negative 5 PSI where a positive 10 should have been visible. Not sure what the measure point is, but it’s where he connected his combo gauge. It uses R134A refrigerant.

I’ve been monitoring the empty unit for the last two days. A graph of the time/temp curve is here.

I warmed the interior to room temp the night before, and I’ve been running it since, with only two door openings, clearly visible in the graph. This is with the temp setting at 5 on a 1-10 scale (i.e., mid scale).

Here’s the question:

It appears to be regulating, but the hysteresis seems excessive, and its cooling rate is too low. Specs on the sealed system indicate 280 BTUs, which seems small if it means 280 BTUs/hr.

A long time ago, I wrote software for a cryo freezer (-190C!) using combo Freon/ammonia refrigerant and recall something called “de-logging”, which was a 10 minute break in compressor operation per 8 hours of constant operation during pull down. I am sure this Subzero doesn’t have a de-logging feature. It was to prevent compressor oil from gelling. The tech on my unit blamed oil as the culprit in the blocked evaporator.

I need to build the case that the unit is broken because it ran excessively without a delog feature, traceable to the power outage and resulting cool down demand. I also would love to know if there is any way to clear an oil clog in the evaporator without replacement. I have access to the entire evaporator surface (from the front), and the unit is obviously sufficiently charged.)

Any/all advice info greatly appreciated. I am an electrical engineer and know enough about refrigeration to get frostbite. Thanks in advance!

(Note: I posted this on a refrigeration engineering site, too)
posted by FauxScot to Technology (7 answers total) 1 user marked this as a favorite
I need to build the case that the unit is broken because it ran excessively without a delog feature, traceable to the power outage and resulting cool down demand.

Maybe I'm missing something, but are you suggesting that the fridge failed because it wasn't already cold when it started up after the power outage? If that were the case, wouldn't you expect new refrigerators to fail the first time they were plugged in?
posted by jon1270 at 7:26 AM on September 8, 2011

hey jon...

yup, i'd expect that, too, except perhaps a new fridge might not be loaded with a large, room temperature payload to cool, and thus, might run less. Also, it might be that my graph shows a poorly performing system, which needs to run longer than normally needed to get to the setpoint. That's why the graph is there. Is that slope too shallow for a $10,000 refrigerator? These things I do not know!
posted by FauxScot at 7:35 AM on September 8, 2011

"The tech showed me a pressure reading of negative 5 PSI where a positive 10 should have been visible. Not sure what the measure point is, but it’s where he connected his combo gauge. It uses R134A refrigerant."

These numbers are of limited use without knowing the high side pressure and without being able to observe frost patterns.

"I need to build the case that the unit is broken because it ran excessively without a delog feature, traceable to the power outage and resulting cool down demand."

I'm not aware of any residential or small commercial refrigerator with a "delog", a term I'm not familiar with, feature with the caveat that systems with defrosters have "delog" as a side effect of the elements coming on periodically. And pretty well all residential refrigerators will run continuously without damaging the refrigeration system. It's expected that a fridge will be called upon to cool it's entire contents from room temperature at some point in it's life; the grid fails often enough for this to be a design specification. The food wouldn't be safe to eat but the fridge would be fine. Even long running periods at excessive head pressures caused by a low side blockage aren't a serious problem because the inherent self limiting nature of the hermetic system designed to function without accumulators.

I've never heard of a clogged evaporator caused by oil break down but it's possible i guess. There are refrigerant/oil incompatibilities that can lead to system failures however they aren't related to run time generally. The fact that sub-zero is replacing the evaporator out of warranty at no charge to you probably means they've identified a recurring problem with these units worthy of a silent recall. Whether the failure was a coincidence or related is probably only something Sub-Zero would know via empirical evidence.

Finally, and be aware this is a WAG as I haven't seen your fridge, it's probably not the evaporator itself that has clogged rather the high pressure line (called a capillarry tube) leading to the evaporator which generally has an inside diameter a few hundredths of an inch in diameter. Much easier to plug then then the 3/8ths or so evaporator coil. The evaporator and cap tube are commonly shipped as a factory assembled unit and the tech will refer to the assembly as the evaporator because that is a term many homeowners have heard before. I've got the tools (a high pressure pump and assorted pigs) to clean the line but I've used it only a handful of times as it is generally cheaper and more reliable to just replace the tube then it is to clean it.

PS: just because your ice didn't melt doesn't mean your frozen food is safe. Ice doesn't melt till zero degrees. Lots of foods, especially those with high sugar contents, will thaw well before ice melts.
posted by Mitheral at 10:01 AM on September 8, 2011

Thanks for that great info, Mitheral. That all makes sense. (I am a little mind frazzled today.)

Wondering if you saw the graph and what you think about the cool rate and the hysteresis? Looks pretty lame to me, but I have no comparisons to view. Also, there's a little peak at the turn on point for the cooling phase. Any physics explanation for that? Seems quite repeatable.

I am about to get into the remaining technical details of the entire system, and I'll have more info soon. The unit is a hermetically sealed system, and the manifold gauges the tech used were connected to what appears to be a stem valve refrigerant input point.

Incidentally, the other spec that I coded into my cryo freezer was a minimum off time for the compressor. It was several minutes, in fact. The client did not want the units to cycle quickly and they had me put in this delay. I can look at the source code and see what I coded, but I don't think it was adjustable.

I was at the auto parts store earlier and they sell R134A to the general public, as well as manifolds and valves and gauges to determine connect the cartridges. Oddly, I thought you had to be 'certified' somehow to buy refrigerant? Am I thinking about only R12 or other ozone depletors?
posted by FauxScot at 12:14 PM on September 8, 2011

As I'm sure you are well aware, prolonged operation at lower than rated voltages can wreak havoc with induction motors because it can cause them to draw enough current to burn the windings. (Here is an interesting 2004 article on the coming wave of brushless DC motors for refrigerator compressors--which I am assuming your fridge is too old too have a chance of using.)

Doubtless your fridge has breaker protection against this, but breakers aren't perfect, and operation at a voltage just high enough not to trip one (or too low to trip one, conceivably) could possibly stress an older compressor such as yours to a point of damaging it.

It would be interesting to know whether your power spent any considerable time in brown-out mode before going out altogether, or, I'd guess more likely, as it came back on.

Have you been monitoring line voltage, by the way?
posted by jamjam at 2:29 PM on September 8, 2011


I haven't been monitoring line voltage, but the compressor shows no signs of failure, so I presume that it didn't operate that long at brownout voltages. I have been monitoring the box temp for several days now; 2 unloaded and overnight with the contents back in the fridge. Still cooling, but wow, the cooling rate is crazy low... which makes sense since the sealed unit has but a 280 BTU/hr capacity, but man, slow.

This unit doesn't appear to use brushless DC motors, either.
posted by FauxScot at 4:18 AM on September 9, 2011

leak in evaporator, very slow. most freon gone.

tech replaced evaporator and it's cooling fairly quickly now.

Unstable operation was due to only having about 1 oz. R134a in a system intended to have 8 oz.

Second set of operational graphs being collected just for general yuks. I am also adding a small paperless recorder/alarm to the fridge (Omega RD8250. Sweet! I happen to have 3 available so one of them is going in my Subzero as a payload spoil alarm and door ajar indicator.)
posted by FauxScot at 2:40 PM on September 15, 2011

« Older Is it cost-effective (affordab...   |  My company got lucky and we're... Newer »
This thread is closed to new comments.