What makes compressors fail

[geoexchangeman]

What is the number one cause for compressor failure.

[Wild Leg]

Human error.

[Shophound]

Compressors don't die; they're murdered.

Number One cause of compressor failure: improper installation procedures. This encompasses a long list of things: improper evacuation, not sweeping the lineset with nitro during brazing, no liquid line drier, ductwork improperly sized, system improperly sized, OVERCHARGE, improper piping procedures, etc.

Runner up is improper maintenance, even on a properly installed and sized system. Dirty coils, poor service procedures, etc. will put a compressor on death row as quick as anything.

[Toolpusher]

Electricity being applied,

They don't fail till they are energized.

[bornriding]

For me, the number one cause of existing compressor failure is low freon levels.
2nd cause would be liquid getting to compressor.

Wonder why anyone would think to answer about a new install ?? Only had two compressors in thirteen years fail that we installed. And one of those was in a low-rent apartment complex.The other was a 12 Seer system.

[Shophound]

Wonder why anyone would think to answer about a new install ?? Only had two compressors in thirteen years fail that we installed. And one of those was in a low-rent apartment complex.The other was a 12 Seer system.

I must commend you on your low failure rate in systems you've installed. Sounds to me like you're doing things right if that's the case. :p

Why did I respond as I did regarding installs? I didn't just have new installs in mind, but repair/replacement efforts as well. If a system isn't sized and installed right, the compressor will pay for it, sooner or later. Bad repair efforts also can jeopardize a compressor's life. Once I bothered to investigate why compressors fail, I have yet to find one that's the fault of the manufacturer.

I also think overcharged systems are a larger culprit in compressor failures than undercharged, as an undercharged system is more likely to see the compressor kick out on IP, even under lighter loads, which would draw attention to itself faster (an undercharge usually means there's a leak, which also calls attention to the system). An overcharged system might kick out only under extreme conditions but otherwise run a greater percentage of time under potentially floodback conditions, which is anathema to valves and lubrication of a compressor.

Electrical failures account for another percentage of comp failures, ranging from single phasing of three phase compressors, to inadequate voltage supplied to a unit, to dirty contactor points, etc. The latter two I mentioned might be a result of installation or maintenance effots. Undersized wiring and no regular inspection of electrical components would contribute to a failure.

[Edited by shophound on 03-16-2005 at 04:39 PM]

[bornriding]

Did not mean anything offensive !
Have just heard a lot of threads that indicate compressor falures on fairly new systems.
And I did not think of it the way you did, as i don't see many failures that I can blame on the system. Although there are a few.
MOst times, around here, if the freon is low, the HO does not realize anything except he is not cooling, and then some time later decides to call for repair. By the time he has called and I arrive, the unit has been on a long time ( maybe even over the weekend ), the compressor has overheated & shut off, cooled down & run again, several times before I get there. I do try to remember to tell them to shut it off, but sometimes I forget.
And around here, filters & coils get real dirty, but the HO doesn't know it until his cooling has stopped. By that time, liquid has gotten to the comprssor more than one time, before he calls it in.

[airworx]

flooding - caused by small return air, overcharge etc..,

[tinknocker service tech]

compressors normaly dont burn out
human error is the bigest cause for failed copressor

[Shophound]

Shop
Did not mean anything offensive !



None taken. Just wanted to clarify.

We're cool. :cool:

You have a good point about homeowners being slow to call in a unit that's undercharged and has a compressor that's cycling on IP. I work on commercial stuff that I'm directly responsible for so I tend to forget the homeowner factor when I post! :D

Your last paragraph falls within my "runner up" catagory of poor maintenance. So, I guess we could say that perhaps a leading cause of compressor failures in the residential sector is, well, homeowners! :D

[bornriding]

Residentially speaking: You are absolutley right at least in this area. Most compressor failures that I see are HO induced.Of course a lot of these are rentals or apartments.
Done by tenant ( or not done as the case may be )

Look forward to bantering with you again !

Richard

[daytonafan]

#1 Cause of Comp. Failure.
Here's my story: 1. Luxury Apartment Complex
2. 128 Apartments
3. 4 Bozo installers
4. 1 mildly warm Indiana summer
5. 24 (yes, 24) comp. replacements
6. 1 extremely mad boss


One co-worker actually said "Hmmm, must have been a bad batch of compressors at the factory." We have a new company policy about installing air conditioners now. So sad it took something like this to make it come about.

[RoBoTeq]

Originally posted by shophound
Compressors don't die; they're murdered.


I like that :D

Compressor failures within the first year are usually caused by improper installation, application or just plain old bad work habits.

Compressors that go for more then a year are usually killed by overcharging. Most overcharging is due to improper diagnosis of symptoms. Most units that are grossly overcharged are done by techs who only read suction line data. Low suction caused by low indoor air flow due to dirty filters and/or coils are misconstrued as low refrigerant. Adding refrigerant brings the superheat of a system with low indoor airflow back to normal and allows the system to operate but with a reduced capacity and with possible slugging of liquid back to the compressor.

[hvacbear]

Bornriding and shophound read my mind and robo filled in any voids. Other than their reasons I have seen a few compressors fail due to improper 3phase power. For no apparent reason a comp failed and the voltaged were something like 221/222/174 we made a big fiasco and the electricians came in to swap some loads around and do a reassesment.

[comfortmaster1]

Originally posted by geoexchangeman
What is the number one cause for compressor failure.

When they don't study for their test.

What's up with all the credentials?

[docholiday]

What makes compressors fail ?

One Word....







Stress.

Take your pick, Thermal Stress, Electrical Stress, Mechanical Stress, and Chemical Stress.

As Shophound said, they are murdered.

[schmuck]

Originally posted by docholiday
What makes compressors fail ?

One Word....







Stress.

Take your pick, Thermal Stress, Electrical Stress, Mechanical Stress, and Chemical Stress.

As Shophound said, they are murdered.


I like that answer.

[geoexchangeman]

All these answers are good. The things that make high compressor ratios are the greatest killers. High compressor ratios are the number one killers says the compressor manufactures. Next in line is the wrong voltage either high or low. Flood back of systems with fixed bore metering devices is also up there from over charge. Refrigeration low back pressure unit are made to run with higher compessor ratios. But A/C Units are made to run with high back pressure and therefore lower compressor ratios. So if they are installed and have to operate with less than design air flow they will hit the graveyard quick. I can't leave with out saying a word about Geoexchange. They run with compressor Ratios of less than 2.5 to 1 in cooling some time as high as 3.5 to 1 in heating and have high and low pressure protection. Air source HP unit have to operate with compession ratios of 6 to 1 under normal conditions and let alone the conditions they get in when the HO don't keep the filters and coils clean. Add the defrost cycle, I don't see how the last as long as they do no matter what SEER rating they are. The low end models have only the internal overload to protect them. As I have said geo units love what they are doing and keep on doing it for many years after the warrenty is gone if the HO will just change the air filter. If not the unit will trip out on high or low pressure and must be reset because of the lock out relay. Is that a plus for Geoexchange.

[docholiday]

Originally posted by RoBoTeq

Originally posted by shophound
Compressors don't die; they're murdered.


Adding refrigerant brings the superheat of a system with low indoor airflow back to normal and allows the system to operate but with a reduced capacity and with possible slugging of liquid back to the compressor.

I think what you meant to say was it brings the suction pressure up to normal but the superheat will go down if you have low airflow and add refrigerant, which will lead to slugging. This is prevented by actually reading superheat and not charging by pressures alone.

[RoBoTeq]

Originally posted by docholiday

Originally posted by RoBoTeq

Originally posted by shophound
Compressors don't die; they're murdered.


Adding refrigerant brings the superheat of a system with low indoor airflow back to normal and allows the system to operate but with a reduced capacity and with possible slugging of liquid back to the compressor.

I think what you meant to say was it brings the suction pressure up to normal but the superheat will go down if you have low airflow and add refrigerant, which will lead to slugging. This is prevented by actually reading superheat and not charging by pressures alone.

In most cases this is correct. I was thinking of one particular instance where the technician made certain that the superheat was correct but ignored the fact that the suction temperature was on 2ºF. In this case, the suction pressure was also low but this tech stuck to his guns and make sure he had the correct superheat. Damnest thing I've seen. The unit was actually operational with solid ice all the way back to and halfway up the compressor housing.

[Shophound]

Air source HP unit have to operate with compession ratios of 6 to 1 under normal conditions and let alone the conditions they get in when the HO don't keep the filters and coils clean.

Hmm...let's crunch some numbers from the performance curves for a Trane TWA060C400A1 heat pump.

Heating mode, conditions as follows: OAT = 30°F; RAT = 70°F db; yields a condensing pressure of 180 psig (95°F SCT) and an evaporating pressure of 36 psig (15°F SVT), according to curves. Add 14.7 atmospheric for absolute pressures of 194.7 and 50.7 psia, respectively. 194.7/50.7 = 3.8

To approach a compression ratio under given conditions in heating mode to near 6, the head pressure would need to be boosted to at least 220 psig and the suction drop to 26 psig (yields a CR of 5.8). At this point your SCT would be ~109°F and SVT ~4°F, which means the indoor coil is ~40°F over RAT and the outdoor coil is ~26°F below OAT. Doesn't sound like the normal operating parameters the curves are looking for.

I I'll grant you that water and ground source heat pumps have the advantage of a fairly consistent condensing and evaporating medium, hence a lower compression ratio across the board.
However, I can't imagine an OEM of an air source heat pump willfully designing heat mode CR's of 6 or greater.

[MadeinUSA]

There is no way that unit could be operational with lines, and half the compressor frozen. In that circumstance on a R22 high temp system, the coil would have to be frozen solid with no air flow, therefore causing his low suction reading.

With a suction saturated temp of 2 degrees, his suction pressure was 25.6. His suction line temp being frozen would be 32 degrees, so 32 – 2 = 30 degrees super heat. Was this bozo thinking he was charging a car since his low side was at 25. I guarantee you Robo, on the phone he was telling you he had air flow, but you and I both know that’s impossible when you have a solid block of ice for a coil.

Not arguing here, we both know a coil temp of 2 degrees will produce a block of ice with R 22 in a high temp application.

Edit: Heck even in a R22 medium temp application, a 2 degree coil would be frozen on you, so a residential being it is high temp is darn sure going to be frozen solid.


[Edited by madeinusa on 03-17-2005 at 01:36 PM]

[RoBoTeq]

Evap coil was frozen solid and the system was operating. What will really get you is that this was a Goodman 10 SEER five ton system.

I have photos of this system but I do not know how to post them on this forum. I will start a thread elsewhere to show you this system.

Like I stated, I too was amazed to see this unit operational.

[tinknocker service tech]

trane has in its factorie in texas a compressor with a closed loop running for last i heard twenty five years with out ever beening turned off. i is a solid block of ice
and still runs. they say its hard to stop a trane
point being be it installer, service tech. home owner. power. charge. it stills comes down to one thing
human error is the main reason for compressor failure
my opinion only
do it right the first time or keep going back

[docholiday]

As I said.... Stress

[dash]

Originally posted by tinknocker service tech
trane has in its factorie in texas a compressor with a closed loop running for last i heard twenty five years with out ever beening turned off. i is a solid block of ice
and still runs. they say its hard to stop a trane
point being be it installer, service tech. home owner. power. charge. it stills comes down to one thing
human error is the main reason for compressor failure
my opinion only
do it right the first time or keep going back

That compressor,"Snowball",died a year or two ago.

[tinknocker service tech]

dash
realy dam i am sorry to hear that it was about ten years ago when i saw it and read about it again about two years ago guess they need a new one lol
still say human eror it main cause

[RoBoTeq]

From what I understand the only reason "snowball" died was because it was shut down to move and then started back up again.

[docholiday]

Ice has nothing to do with compressors failing, as long as it's not liquid returning, and there isn't too much superheat its ok. The problem is, when the coil ices, it usually results in slugging. It's simple enough to ice a system then adjust the charge so it stays frozen and still doesnt slug.

Many of the commercial systems I used to service were balls of ice but then again they were machines that operated with evap conditions below zero with proper charges.

[Special Ed]

Looked at a Bryant 5ton air-source HP today. Lineset is about 25-30ft long; suction line is 3/4 & the liquid line is 3/8. Compressor is no longer with the living.

How old do you all think it is?

[wasp]

I had a 7 1/2 ton Trane split, totally froze up. Shut it off and thawed, restarted, found superheat at 0 deg and ambient was 70. Txv stem was all the way out, set to 15 degrees superheat and did not have to adjust charge.

Just changed out an 06D, was mechanically failed (no oil in compressor), changed compressor, swapped unloader from old compressor to new compressor, changed drier, pulled vacuum charged and checked operation. Compressor would not unload (failed unloader), but also found thermostat (30-90 deg range) in return air set at 40 degrees. All oil that went to evaporator from old compressor came back to new compressor. Bled off excess oil from oil pump to proper level and then changed oil.

Oh yeah, too much oil will kill one also.

I love going behind other people and finding their screwups.

[AllTemp]

I'd say flooding 1st caused by over charge 2nd poor maintenance and 3rd but not least compressor design...

Look at those Maneurop Pots you can't hardly slug em... The whole dam compressor is oen big accumulator under that can...
http://cc.danfoss.com/SW/Product_presentation/Images/Maneurop/Group.jpg

If everyone would build em tht way there wold be less failures...

Was Snowball built in this manner?

[Wild Leg]

Originally posted by AllTemp
...Was Snowball built in this manner?

Almost an exact copy.

[Wild Leg]

Originally posted by RoBoTeq
From what I understand the only reason "snowball" died was because it was shut down to move and then started back up again.

So, are you saying snowball was murdered?

I wonder if they did an autopsy?

I'll ask around.

I know somebody who knows somebody...

[tinknocker service tech]

looks like trane answered the question
human error
poor snow ball he was killed