Oil return problems

Freezer store-room, installed 1994.
Original situation: 2 independent Copeland Demand-cooling compressor (R22) on 2 KUBA evaporators, electrical defrost. Danfoss TE5.
No oil separator nor suction accumulator.
2 HE’s in the discharge for make-up water for the ground heating.
Since the install no big problems. (broken heater in the evaporator, condenser fan fault,…)

Round March this year, one compressor went in fault. Piston was broken.
Reason??
Client does minor repairs themselves and sometimes when a heater in the evaporator is broken, they wait to repair it till they have 2 or more broken heaters or unit is completely frosted.
Possible cause: ***** washout in the sump??

Installed a new compressor, standard compressor (no demandcooling) and retrofitted to R404a. Blow out the lines (+/- 80 to 100 ft) with OFN. Let it run on POE and replaced after 4 hours running with new POE.
After 1 day, replaced oil again.

Appr. 10 days later, call from clientà New compressor went in fault.
Cause: broken pistons.
No oil in sump.
Searched for fault: tech had installed a new TE5 and installed an orifice 3 instead of a 1.
***** washout.

Installed new compressor. SH within the limits (20 K measured at inlet compressor)
After 1,5 months running, again strange noise (knocking metal noise) in compressor. It’s just like a piston is has too much spacing. Again, almost all the oil has gone out of the sump.
Checking Oil Safety device: wired wrong by ‘someone’

Is oil return so worse with R404a compared with R22?
Speed in the suction lines is approximately the same.
We did retrofits in supermarkets with more then 2000 ft of lines and never had those problems.
Also never had oil return problems with R22, at least no such a problems.

Installing an (coalescent?) oil separator?

mornin Peter! i would check evapps clear , clean, sh, install standard oil separator, sounds like key is unclear(icing) coils ecessive run times low sh and velocity, parralel compressors or independent compressors? if parr. which stage problem?

Hi Peter.

In all the oil problems I had, always after changing the orifice to a smaller one and installing oil separator, problems has gone.

One more thing you can do is making a 1/4 pipe with SV from the lower point at the collector pipe at the evaporator to the suction line operating via a timer controlled by the compressor main contactor.

I have done this on a unit where the evaporator wes 20m below the condensing unit. No oil problems there anymore.

Chemi

Thanks already, just a quick reply while eating


2 independent compressors and evaporator is +/- 4 m above the compressor, lines sloped towards compressor.

Bon appetite Peter.

Still, check the evaporator design and see how the pipes are arranged.
The oil is there somewhere.

Chemi :)

That's true and it only can be in the pipes or evaporator.
But why weren't there any problems during almost 10 years with R22? That's bothering me now.

Hi Peter.

R-12 was the best with oil return. Never had any problems.
R-22 is good but not as good, R-404 is new and now you know that it has problems with oil, but maybe pipes diameter should be changed or the pipes in the evaporator should be arrange different.

I can see that the industry is still struggling with the new refrigerants.

Its something you will have to find out by yourself, very slow.

Then you will be an expert on "oil return in refrigeration systems operating with R-404A"

Chemi :)

With most of these retro-fit problems the fundamental problem (piping design?) is always there, it's just that the R22/mineral oil combination is more forgiving.

Would the quick fix not be to install an oil separator.

Hi Dallan.

Oil separator will not always solve the problem, its the refrigerant velocity that make problems.

From my experience, there are more oil problems with new refrigerants, than with ssshhhhhhhh ...forbiden refrigerants ;)

I think that all the pipe work i.e. condenser, evaporator, all gas lines will be designed in a different way then today to overcome these problems.

Where about do you come from in Scotland?

Chemi

Does the system pump down in the off cycle and defrost cycle? What are the pressure control settings? Is the evaporator trapped (suction line goes above evap and then down to compressor)? Is the defrost terminated by time or temperature?

What is the subcooling? The very first thing I would check for is overcharge.

Indeed, pump down in off cycle and defrost cycle.
Pressure control settings (bar)
LP: cut out at 0,1 bar, Cut in : 1 bar
HP condenser fans: cut in 15 bar out at 14 bar

Suction line is indeed trapped: goes +/- 3 feet up in one size smaller diameter than the actual suction line.

Defrost is terminated on temperature (5°C) and a safety time of 35 minutes, whichever is first, 4 defrost/day. Each time we saw the evaporators, they were free of ice.

Subcooling is +/- 5 K, discharge cooled by the HE installed for the underfloor heating (preventing frost building up in it)

Why do you think on an overcharge Gary?
Goes always in pump down and superheat is each time we measured it at the outlet of the condenser between 8 and 10 K, all depends on he place where you measure it and who did it.

Peter.

PS: Did you heard the news about the terrible gas accident (explosion of the main gaspipe from Belguim to France) in Belgium: more than 15 dead, some were throw in the cornfields, some burned in there cars because they just passed by on the highway nearby, more then 100 injured, more than 35 in serious life danger (don't know the correct word), most of them are police officers and fireman's.

Hi Peter.

Yes, I heard on the news here about that terrible accident, its a shame that innocent people got killed like that.


If the evaporator is trapped, then it behaves as flooded evaporator.
All the oil is there at the bottom.
Now you can do what I suggested you with the 1/4 pipe from the bottom of the evaporator to empty it from the oil while compressor is running.

If a liquid separator is installed, it would be a good idea to check that its not full of oil too.

The little hole for oil return at the bottom of it, gets clogged with dirt sometimes and stop the oil from returning to the sump.

Chemi

How about putting the system on an extended defrost with a high termination, generating a lot of heat. This should wash back most of the oil if as suspected it is logging in the evaporator.

If the oil comes back you could install a Temprite filter type oil separator which works well at low discharge velocities. I have retro-fitted packs with these sepatators and the amount of oil that they reclaim from the system can be really surprising.

With a 10 minute defrost would we get enought heat to keep the coil clear, even with 8 per day.

From experience for the defrost cycle to be effective as an oil harvest is there a minimum amount of heat (lowest termination temp) that needs to be raised before it is effective.

Does anyone ever add a different refrigerant to R404a systems with oil problems, I can just about remenber when it was quite a common occurance to add some R12 to low temperature R502 systems to help with oil return.

The only time I have done it recently is when working on ULT equipment. These units come from the factory with the first stage using R404a (or maybe R507) with some R134a to help the oil round and R508 in the second stage with some R290 to help the oil round in the second stage.

Anyone think it is worth a try adding some R134a to a LT system with oil logging problems. Would it help or hinder performance if it was added to a LT supermarket pack for example.

I just installed 7.5HP scroll running on R404A............last month
I better go check oil sight glass tomorrow

Next time............Some Wise Ass tells us the hole in the sky is getting bigger.............and we need new gasses.

I will personally go shove him in that bloody hole myself!!!!

Or ill go live in Israel.............and see my darling R12...........502 again !!!
How I miss them

Hey Chem............You not worried about competition are you????

Competition is the fuel I'm using. Why? :)

Chemi

This is an interesting problem.

Apparently the system worked adequately for many years, and the evaporator has not been changed, nor the piping.

Having a riser at the coil outlet does not make the coil act as a flooded coil.

The pistons have broken twice, for reasons that are less than satisfying, once with R22 and once with R404A. I would be curious to know if any repairs were done prior to the first compressor problem, and what those repairs were.

There are a lot of R404A systems out there that are not having oil return problems, as far as I know.

I am torn between viewing this as a floodback problem or an oil return problem. Possibly it is both.

We are missing something here.

What temperature is the freezer? Evaporator air in and air out temps might tell us something.

Why would I suspect overcharge? If the solenoid opens, and there is a significant delay in starting the compressor, the entire refrigerant charge can migrate to the coldest point, i.e. the evaporator. When the compressor starts, this can come flooding back. The larger the charge, the higher the danger. However, trapping the coil should prevent this if the system is not grossly overcharged.

Also, if the defrost is terminated by time duration rather than temperature, the fans will start immediately. making the coil very cold very fast, adding to the rapid migration. Usually this is accompanied by heavy frost on the walls, ceiling, and floor, since the fans are blowing out warm moist air. Time durations should be excessive. They are for backup only.

Quote:

---

Originally Posted by Marc O'Brien

Dave, I didn't think R134a had good miscibility? Adding a touch of any low pressure HC does do wonders.

---

I agree. Propane (R290) for example, has pressure/temperature characteristics similar to R22, and a small amount can move a lot of oil.

I know it sounds blasfemish, but a little R-12 in the existing refrigerant brings your oil back, so you have time to think and to engineer a better system.

Quote:

---

Originally Posted by Gary

Apparently the system worked adequately for many years, and the evaporator has not been changed, nor the piping. ^

---

Correct, nothing changed. We did such a rtrofits already many times, never had these problems.

Quote:

---

The pistons have broken twice, for reasons that are less than satisfying, once with R22 and once with R404A. I would be curious to know if any repairs were done prior to the first compressor problem, and what those repairs were.

---

Even better, just opened this afternoon the new compressor and this is what I found:
* Oil quantity: +/- 5,5 l., so this is OK.
* Oil was grey, due to very small aluminium particles.
* 1 of the 4 pistons broken in 20 pieces,
* 2nd has almost 2 mm spacing on the crankshaft and should have broken any time if we hadn't opened it.
* Piston 3 and 4 seems OK.
* Valve plates OK.

I removed the melted aluminium on the crankshaft with HCl and there are almost no scratches on the shaft.
I was surprised that I could remove this so easily.

I also see some small copper color traces (actually more spots). Copperplating?

Quote:

---

There are a lot of R404A systems out there that are not having oil return problems, as far as I know.

---

That's also why it's bothering me.

Quote:

---

I am torn between viewing this as a floodback problem or an oil return problem. Possibly it is both.

---

Due to the fact that oil quantity is OK, it's more a floodback problem.
Saw also this morning when we removed the compressor that client had installed an additional time delayed contact between the SV so that they can shut down the compressor for some minutes when they're in the freezer to work.

Quote:

---

We are missing something here.

---

Sure, I'm also guessing what the cause can be for all this problems and that's also why I'm asking you, the pro's, for advice.

Quote:

---

What temperature is the freezer? Evaporator air in and air out temps might tell us something.

---

Temperature in the freezer is set at -20°C. Will measure air in/out probably beginning next week.
The last time, we measured subcooling and superheat, LP and HP, AMPS on the 3 lines, oil pressure because we needed to fill in this figures on the warranty papers.

Quote:

---

If the solenoid opens, and there is a significant delay in starting the compressor, the entire refrigerant charge can migrate to the coldest point, i.e. the evaporator. When the compressor starts, this can come flooding back. The larger the charge, the higher the danger.

---

SV opens when thermostat asks for cooling, and compressor starts then in less then 5 seconds.

Quote:

---

Also, if the defrost is terminated by time duration rather than temperature, the fans will start immediately.

---

Defrost is stopped on temperature (+8°C) and additionally with a safety time delay (45 minutes)
A normal defrosting takes +/- 25 to 30 minutes.
But,... the fans can not start as long as the evaporator coil senses temperatures higher than -5°C.

Quote:

---

Originally Posted by Marc O'Brien

Is there any possibility the compressor is short cycling?
What are the wear patterns, on the crank, big end and small end?
Any signs of wearing surfaces overheating?
Photo's of these parts will help a lot.

Also, consider this scenario: http://www.fridgetech.com/discussion...es/42/128.html

---

Found sufficient oil quantity.

Will read the thread on your forum. Seems very interesting.

We wire it like in the 2nd schematic but never install a reset knob.

In the heater lines, a compressor NC contact so that compressor must first shut down before they can start to heat.

Most probably cause: we had a ***** washout but why did it happen???
Have to recheck everything: the complete defrost cycle, pressostat settings, door contact, orifice in the TE5, ...

There is one EWPC 1000 for both compressors, 1st is the 'master' evaporator and the 2nd defrost is terminated via a mechanical thermostat (set on 10°C).
Perhaps the master senses to fast a defrost end whereby the 2nd is not complete defrosted with ice building up (??!!)

Both have additionally a mechanical thermostat as a safety cut-out, wired in line with the heaters.

Quote:

---

Originally Posted by KosmisB

I know it sounds blasfemish, but a little R-12 in the existing refrigerant brings your oil back, so you have time to think and to engineer a better system.

---

I am not entirely opposed to this, and I know for a fact that it works. I once used a small amount of R12 in a cascade low stage to compensate for a non-functioning oil separator out in the boonies, because I couldn't get a new one right away. The temps ran a little higher, but were steady, and it saved several years of research material inside the lab freezer. However...

The pressure/temperature for R12 is considerably lower than for R404A, thus your P/T chart will no longer be accurate for that system, given substantial amounts of R12. This could also effect TEV operation. R290 has characteristics much closer to that of R404A, and is superior in its oil carrying capabilities.

Either way, only a small amount is needed, or should be used.

Quote:

---

Originally Posted by Peter_1

Saw also this morning when we removed the compressor that client had installed an additional time delayed contact between the SV so that they can shut down the compressor for some minutes when they're in the freezer to work.

---

Are you sure the control shuts off the compressor, not the fans? If they are not shutting down the solenoid, this could be a compressor killer.

You may want to see if the compressor floods when that timer is energized and/or just after it times out. What happens if this timer is energized during defrost? This control merits very careful scrutiny.

I would like to see a complete and accurate wiring diagram if at all possible. The devil is in the details.

Indeed, it were the 2 pistons on the opposite site of the oilpump.

The time-delay contact relay shuts off the SV and the compressor goes then in pump down.

Don't you think a 1 bar diff on the LP cut out is too low. If there is no anti-cycle timer after a defrost if the fan delay time is too long it is possible that the compressor could short cycle.

Hi Peter.

From all the ideas that came up on this thread, it looks as the real problem is flood-back and oil wash out.

A suction line accumulator will add some protection to the compressor.

a coalescing oil separator should solve the problem of the oil.

I would check all the refrigerant control components in the system for perfect operation and the electrical controls and timers for the right operation sequence.

I dont think adding some other refrigerant to help oil return would help to save the compressor in this case.

Chemi

Quote:

---

Originally Posted by dallan

Don't you think a 1 bar diff on the LP cut out is too low. If there is no anti-cycle timer after a defrost if the fan delay time is too long it is possible that the compressor could short cycle.

---

Cut out a t 0.2 bar (-43°C), cut in at 1.2 bar(-29 °C)
The longer you wait, the bigger the chance for overflood.
short cycles aren't the reason for a ***** washout. For some reason, liquid is coming to th compressor.

Marc, can you explain a little bit more the difference of the damaging of the pistons, those situated nearby the pump and those situated nearby the motor windings?
Can i state that if it's liquid slugging, then the liquid first meets the pistons nearby the motor windings and wash away there the oil?
When will you have a broken piston nearby the oilpump?

It was indeed likeyou said.

Hi Peter :)
The problem is ofcourse oil wash out by liquid flood back :(
but why :confused: :confused: probably something to do with the customer fit timer and the compressor being asked to pump down and restart numerous times a day, possibly pumping without the evaporator fan running.
Also I would check the pipe sizes on the suction, even though the oil return from the suction seems ok.
Just a thought, if you wanted to make a compressor send oil around the system what better way than pumping it down a few dozen times an hour.

Kind Regards. Andy. :)

Timer is installed almost 7 years ago.
We have there 2 identical systems, installed besides each other, with equal pipe length, the only difference is that the 2nd compressor is still running on R22.

Timer is as far as I saw it on the first sight connected in line with the SV. If the coil is under -5°C , then the fans are running. Fans are only shut down in defrost and thereafter if the coil is higher then -5°C.

I think that indeed the door opens quite often in a day but this is also true for the other R22 compressor.

I will do a new calculation but oil level in the compressor was OK. It's of course also possible that he oil came back very slowly along the sloped pipes during the days the techs had shut down the compressor but that there was initially (while running) not enough oil in the sump.
Then it makes also sense what you described, pumping the oil in the low side of the system without proper return due to a high on/off cycling.
But, .. the R22 compressor seems not having problems with this.
I will certainly check the whole timer system and the defrost.

Again, a full and accurate wiring diagram would be very helpful. There is indeed a long list of possibilities here. Anti-recycle timing would be a prime suspect if it does not close the solenoid.

It should also be kept in mind that floodback (for whatever reason) is limited by the amount of refrigerant in the system. During the floodback period, this system has enough total refrigerant charge to completely flood the (trapped) evaporator, spill over into the suction line, and fill the compressor with liquid. Does the system really need that much refrigerant? I think not. The evaporator is trapped. This should be sufficient to contain the entire refrigerant charge, without spilling over into the suction line.

For the past couple decades I have been removing more refrigerant from systems than I put into systems. That's how common overcharging is. Everybody does it.

The condenser outlet subcooling would seem to indicate that the system is not overcharged, but is this true? The condenser is probably sized properly for the compressor, but we also have a discharge line HX.

With the condenser capacity and the HX capacity combined, the SCT could very well be close to ambient temp, and it is not possible for the subcooling to exceed the difference between SCT and ambient, regardless of charge. In order for the subcooling to tell us what we need to know, the SCT must be raised well above ambient temp.

As usual, a full list of temperature measurements tells the story. You are checking subcooling and superheat, and that's a very good thing, but given a full set of raw numbers, we can combine them in a variety of ways, each of which tells us something. Differerence between SCT and ambient is just one example.

I agree, Marc. Overcharge is not the cause of the floodback, however I suspect that it is ALSO overcharged. The amount of charge can be the difference between moderate flooding and severe damage.

Hi :) What about R22, the only commonly ***** refrigerant that will carry large quanities of water around without dying. Say R404a was added without washing the system and pulling a long deep vaccumn, could contamination be the problem :confused: POE oil would wash this into the compressor quite quickly, with a faulty oil fail it wouldn't take long for the compressor to fail :eek:
Kind Regards. Andy.

Are the crankcase heaters working ?

Yup, be even then, system goes in pump down, so if there should be a ***** migration, pressure will rise and compressor will pump down again.

Peter,
I’m in agreement with those who believe that you are experiencing flood back so I believe that as a protection against further compressor failure you might want to install a suction line accumulator to add an extra safety precaution. If you decide to do this you might keep in mind that the accumulator should have adequate liquid holding capacity of at least more than 50 percent of the entire system charge and with out adding excessive pressure drop to the system. In addition, to be 100% safe you need 100 percent of the entire system’s refrigerant charge in severe flooding cases. :)