Losing level in oil separator

[lifshor]

This is a propane refrig unit: oil separator is located downstream from Howden screw compressor. Unit will be running and suddenly loses oil level; we've not been able to ascertain cause; maybe synthetic lube oil temp is too cool and propane is not flashing off in separator. Every once in a while we'll blow the filter coalescer inside the separator. Is it possible that the hole on the filter hold-up plate inside the separator is too small? Has anyone experienced similar problems? Possible solutions?

[Emmett]

What type of seperator? If you are blowing a hole in the filter media then A.) you run the risk of plugging the hole
and
B.) you will loose oil, much more quicly then you might think.
I do not think the size of the hole would have any bearing on the fact that you are blowing a hole in the media, this is a function of pressure differential across the media whic occurs as it becomes plugged. Is this a new system? were proper refrigeration brazing practices used? You may also want to investigate whether you are using the proper micron range for the filter media or not, it may be to small. let us know what you find out.

[lifshor]

this is a vertical separator. the compressor discharge comes in on the side of the vessel below the filter support plate. The filter sits on top of the hole in the middle of the plate so the gas goes from the inside out from the bottom up. Gas then comes out the opposite side of inlet nozzle.

When we blow a filter cartridge, it normally fails along the sealing surfaces, either on top or bottom. No hole is blown in the filter, however the pleating of the filter media is unfolded close to where the cartridge failed.

This unit was not new, but we've had problems with it since it was put in service. The pressure differential across the filter is 0.6 to 1 psi and this number varies constantly and quickly. I'm curious if the change in the differential reading is due to the propane flashing off the lube oil. The lube oil is synthetic and has little miscibility with the propane so that as the temperature increases, the propane should flash off.

[RANGER1]

lifshor ,
What is your oil temp & saturated discharge temperature ?

Oil could be foaming or blowing out of seperator, does the oil show up again in seperator if you shut it down or anything?

[RANGER1]

To stop slipping maybe you can instal some type of locator in clamping plates .
If you can ensure propane has flashed off with higher oil temps & still have pressure drop maybe consider bigger coalescers or extra pleats in filter for more surface area .

[lifshor]

lube oil inlet is 120F (48.8C), propane inlet is -10F(-23.3C) Discharge pressure is 210 psig and the temperature 220F (104.4C)

If it were be foaming, but how can I determine this? When there is a trip due to low oil level, after a few minutes (5-10 min.) the oil does return, not to the original level, but a lot returns.

I still think that the size of the hole in the middle of the support plate is too small. It behaves as a flow orifice (venturi effect), so that if the velocity is high enough, it will entrain the oil out of the bottom of the separator, causing the filter cartridge to fail. It probably behaves as a fast moving glob of oil hitting the cartridge.

[RANGER1]

lifshor ,
I probably should have asked for discharge temp as well which I'd expect to be around 60-65 deg C for propane hopefully not 104 deg C .


If oil is suspected to be foaming it may be reduced by increasing oil temp if you can or lowering condensing pressure.
If discharge temp & oil temp are to close together oil will become saturated with propane & flash off .
These type of oil seperator are usually 3 stage ie cyclonic , demister pad & last stage is coalescer .
I would also suggest that you have to have some sort of pressure drop across filter & 1 psi is not much .
Try to increase oil temp by a least 5 deg C & see what happens .

[Magoo]

Possibly the compressor loading rate is too fast, that will increase velocities through separator and carry the oil out of separator.
So monitor system and manually load compressor and figure out a suitable load rate.

[lifshor]

Ranger1: thanks for the tips. One question though, since this screw compressor is an oil flooded compressor, and the oil is the heat sink, I would expect the gas to have a higher temperature than the oil. The temperature I reported above was the propane discharge temperature. I do not know at this point what the oil temperature is at the compressor discharge. I will find out later this am.

In talking to Howden, it seems they want a higher inlet lube oil temp, from 120F to 140F. We have a synthetic lube and propane is not too miscible in it. I guess the higher temperature in the lube oil will cause the propane to flash off and reduce?? entrainment??.

Magoo: thanks also. The problems with the oil level are not restricted to startup. We start it up on manual control and can overcome this problem. However, when the ambient temperature begins to cool, or gets colder such as evenings and early mornings respectively, the loss of oil level becomes more prevalent. I wonder if the oil temperature control valve is working properly?? Thanks!!!!

[spwright]

In my experience these larger machines do not behave well with unstable condensing conditions, the result of changes to the condensing condition however small will cause the refrigerant to separate from the oil, essentially producing foam, which reduces the internal volume of the separator and then carry over through the coalescor section occurs. My advice would be to stabilize the condenser as best as you can. The cause of the oil temperature becoming higher than the discharge temperature is most likely an instrument fault or calibaration error, the must always be at a lower temperature where else does the heat energy come from?...

[lifshor]

spwright: sorry if I confused you. The gas discharge temperature will be higher than the lube oil temperature coming out the compressor. The gas temperature rises due to the heat of compression whilst the lube oil temp increases due to the heat transfer from the gas. The lube oil is the heat sink.

We would like for the refrigerant propane to separate from the lube oil. I understand from Howden that the lube oil temp into the compressor needs to be 140F. At 140F on the inlet, the propane discharge temp will be higher causing the propane to flash off the lube oil in the oil separator because the propane is not readily soluble in lube oil.

So, will a higher discharge temperature improve or exacerbate the loss of oil level in the oil separator?

[RANGER1]

Oil temp has to be higher than the saturated discharge temp approx 10-15 deg C so it will boil out of propane .

Compressor discharge temp will be higher than saturated discharge temp & each application & refrigerant type will have different theoretical temp .

eg ammonia 1250 kpa = 35 deg C saturated
Oil temp 50 deg C = 15 deg C above SDT
Discharge temp out of compressor approx 70-80 deg C
If oil temp is reduced to 35 deg C there is high possibility of oil saturated with ammonia .
Hope that makes it clearer .
Note an oil cooler on a screw is also part of condenser as its transfering heat

[sterl]

Where is your condenser with respect to the oil separator? Where it the discharge check valve located with respect to the oil separtor? How is your oil cooled? Is there an economizer arrangement on this package?

Your oil temp is definitely low and it is very possible that the oil deck return is too small; if this package was originally running on mineral oil the viscosity at high temp may be much lower than that of a synthetic at the same temp, even though they are relatively the same at or near the injection temperature. The same condition will also create a ticker film on the coalescing elements.

But don't start re-working the internals of the oil sep without pursuing the other topics: I presume this is not happeining with the machine off; that we don't have a suction or economizer check valve taking permitting the compressor to back spin and pump oil to the low side?

Is this an NG pipeline or is your propane all propane? If you have a partial presure of some other volatile, you can get some very strange issues with oil though its usually the tip of the iceberg.

Check motor currents and condenser return temperatures....Very possible your condenser is filling up with a volatile to the point where it eventually loses its seal and the volatile "blows over". This will show up as a large difference between the condensate drain temperature and the saturated condensing temperature corresponding to your discharge pressure.

[lifshor]

The propane condenser is located downstream of the separator. There is no discharge check valve. There is a check valve in the suction line.
The lube oil is cooled in an oil cooler, downstream of the oil circulating pump. There is temp control loop that diverts oil around the cooler to maintain the oil inlet temperature at 140F as recommended by Howden.
The lube oil is synthetic and propane has low solubility.
We lose oil level in the separator when the unit is running. If the unit is down, there is no loss of level.
The propane used is not refrigeration grade because there is none available from the suppliers in the area.
We do not have any information on the volatiles to comment on the condenser blowing over. I would expect the non-condensibles to travel to the accumulator.

[RANGER1]

lifshor have you increased oil temp & are still throwing oil over?
Are you 100% sure coalescers are i perfect condition?

Suggest change them & start fresh

[lifshor]

we increased the lube oil circulating temperature from 120F to 140F. The oil level still remains a problem. We have not been able to ascertain the causes leading to the loss in level.

What is the recommended linear velocity of the propane/oil through the filter coalescer inside the oil separator to ensure no entrainment?

The loss in oil level happens randomly, even when we just replaced a new filter cartridge. However, now that we have reduced the gain on the discharge pressure controller and the valves changes are very smooth, very frequently now, we do not blow a filter cartridge

[RANGER1]

lifshor,
Sounds a difficult problem , have you had contact with oil supplier with your difficulties ?
everyone recommends CPI as good source of .
How big is coalescer and what size/type of compressor do you have?

[lifshor]

who is CPI?
We have a Howden compressor, model WRV screw compressor. There is a PLC controlling some loops and logging data. Which variables would you consider essential to track the operation of the compressor? We have Tin, Pin, Tout, Pout, Delta P across the filter, Lube oil Tin & Pin. Any other suggestions?
I looked at the oil separator internals, and the inlet nozzle points down. How much space is recommended between the inlet nozzle, the oil level below? How about between the nozzle and the filter coalescer above? What would be the recommended linear velocity?

How critical is it to have no liquid on the inlet in an oil flooded screw compressor? Howden tells me that if I have oil coming in with the propane, it is not damaging to the compressor because this is an oil flooded compressor. However, they would prefer not to have liquid propane come in.

[RANGER1]

CPI is from Michigan .

The only unit I worked on had a type of gear oil , don't know details.

Linear velocity I don't know.

What size WRV Howden & what dimentions of filter .

Liquid entering machine is not good as it could wash out oil if to much & you also have no control .
Maybe level control system in evaporator is full of oil!

Run oil seperator oil level to absolute minimum & see what happens

Did you find out actual discharge temp on discharge of compressor?

If you remove coalescer can you actually see oil level?
Usually you would see demister pad ( stainless steel mesh).

[Magoo]

I agree with Ranger, the oil temp should be 10>15'C above discharge press/temp., from memory the oil pressure in aHowden WRV should be 35psig above discharge pressure any high and over injection happens which agrivates oil carryover from separator, particually at part load. Screws can handle saturated suction gases probably down to 2.0'C suction superheating, liquid slugs not.
Is separator horizontal or vertical, and is there a stop check valve after oil separator.
I had a similar problem with a 204 Howden and ended up being a worn tephlon seal in the injection tube where the slide valve meets tube.

[RANGER1]

I suspect that Howden on propane has no oil injection at all .
As stated in an earlier post compressor discharge temp should be 60-65 deg C

[Magoo]

Hi Ranger1.
not familiar with propane and screws at all, [apart from the bbq] so what creates the compression seal without oil injection, are the rotors flapping around with no oil. Share your experiences please.

[lifshor]

a few minutes ago, I checked the discharge P and T: 212 psig, 172F (78C)
The size of the compressor driver is 500HP.
The diameter of the filter is misleading since it sits atop a hole that is probably 6 inches ID. Length is about 18 inches, pleated with a wire mesh on the inside and outside.
I personally think that the hole on the filter suppor plate inside the oil separator, because of its small diameter, causes the propane to speed up too fast thereby entraining the lube oil and losing the oil level in the separator.
Anyone know what is the recommended velocity for a propane/oil mix to go through a filter to effect coalescing?
Magoo: wikipedia has a good description of a screw compressor.

[RANGER1]

lifshor ,
500hp could be WRV 255 of some description which is decent size .
Have never seen 1 coalescer on a machine that size normally 2 to 5 depending .
Suggest talk to a coalescer manufacturer for flow rates etc or a contractor in this field .
Is the plant running at same conditions that it was designed for?

[RANGER1]

Hi Ranger1.
not familiar with propane and screws at all, [apart from the bbq] so what creates the compression seal without oil injection, are the rotors flapping around with no oil. Share your experiences please.

Magoo ,
Have seen it on some process applications which were like booster operation .
No oil injection at all .
A mycom booster can to if you want it to as due to lower pressures little internal bypass .

[lifshor]

Ranger1: thank you for all the comments; I'm already talking to a filter-coalescer manufacturer. I agree with you that the size of this vessel is too small.
As to the conditions: it is running as designed, but wth a lot of shutdowns due to loss of oil level in the separator.

[Magoo]

Hi Ranger.
Thanks for comment, I was aware Mycom do alot of petro-chem applications pumping all sorts of nasty vapours around. Do the mains and thrust have isolated lubrication setups?

[RANGER1]

Magoo ,
To be honest I have seen 2 applications where Howden screws are on propane with no oil injection at all .
Overhauled them but never went to site (thank goodness as always in some uncomfortable environment ).
The slide valves didn't even have the oil injection holes drilled into them .
I think propane has pretty low discharge temps anyway .
Looking on Mycom programme they still suggest oil injection .
Howden have big oil flows so maybe they can get away without oil injection .