PDA

View Full Version : Oil lift







Nico
04-01-2013, 12:41 AM
Years ago I have seen a flooded evaporator (brine chiller) on a ship with an oil recovery system using an oil lift.
That came from the Stall or Sabroe stable and it was used to recover oil from the R22 (?) separator.
I think to remember that it was a communicating pipe with the separator
The oil leg was heated (a pipe in pipe heat exchanger with the liquid line?) and at an angle of about 30-45 degree.
Oil supost spilling over in the suction line by force of gravity.
Who knows about them and what design criteria and controls were used (if any).

sandybapat
04-01-2013, 05:42 AM
Nick

I had once seen such arrangement on a Chlorine liquifier leaflet of Stal. In side the shell there was annular parallel plate on sides. The oil normally floats on R22 liquid. The liquid rises in the space between main shell and paller annular plate. This liquid is normally oil reach R22 liquid and this mixture is tapped and taken to that slanted vertical heat exchanger. The liquid fed to Chlorine liquifier was passed on the shell side of this exchanger which is around 40 C and this will evaporate the R22 liquid from the mixture tapped from liquifier shell. The oil will R22 vapours are led to suction line for oil return to compressor.

I will try to locate the leaflet if possible and post it with in couple of days.

Sandybapat

Nico
04-01-2013, 07:43 AM
Thanks Sandybapat. Thats the system I am talking about.
I am curious how the liquid flow from the evaporator or separator to that heat exchanger where refrigerant is boiled off is controlled (if that is the case). Furthermore I would like to know how the flow to the compressor is controlled (if there is any control). Would that work for only one compressor or what if I use it for a group compressors and how should the piping be arranged.

RANGER1
04-01-2013, 10:20 AM
Niko,
The liquid/oil mixture from evaporator flows into liquid line heat exchanger, only when liquid line is feeding evaporator liquid makeup.
So main liquid line makeup has a solenoid valve as well as oil skimmer from evaporator.
They both energize when evaporator requires liquid makeup, as this is where oil skimmer gets its heat from the flowing refrigerant.
The oil skimmer solenoid must be direct acting, so it has no pressure drop.

On multiple compressors the oil skim line is still returned down suction line to whichever machine is running, but all compressor oil seperators return to a common still. The still which is maintained at a slightly higher pressure than suction or compressor sump pressure The still is vented back to suction line through a pressure regulator. Maybe set at several bar above suction pressure for the driving force.
A heater can also be installed in still.

Each compressor has a float in sump & it returns as required from the still through a solenoid valve, which would be energized when that machine is running. Normally oil return solenoid to sump has a delay timer to allow for oil to settle if foamy in crankcase.

Tycho
04-01-2013, 04:15 PM
see my last two posts here http://www.refrigeration-engineer.com/forums/showthread.php?36971-oil-return-in-flooded-evaporators-with-R507

Disregard the tex valve on the liquid inlet for use on a flooded chiller. instead the oil "boiler" needs to be mounted with the bottom inlet from the drop leg, and it should be placed according to the liquid level in the chiller so that it is around 50-60 % filled on the cold side.

the sabroe ones are the same, only they are placed at 45 degrees

Nico
05-01-2013, 12:26 PM
The design with the connection to the pump and with the expansion valve that you refer to makes sence to me Tycho.
Proper control under various conditions, easy to calculate and posibility to distribute to several compressors.
I would do it that way too.:)

However, I like to know about the "troublesome" design where the density of the oil and refrigerant are crusial for proper working of the oil rectifier.
This is what I understood design wise so far (see attached drawing).

I have understood the idea is this (please correct me if I am wrong):
We look at a running system, at least one compressor is running, motor valve V04.02 is open for >15%, solenoid V07.02 is open.
Refrigerant with oil will flow into heat exchanger H07.04 and temperature of the refrigerant/oil mix starts rising.
Refrigerant boils off and that gas returns to the compressor through the piping on top.
The concentration of the refrigerant with oil goes up, until it reaches the maximum miscibility for that refrigerant with that oil at that temperature and that pressure.
If more refrigerant is boiled off than the we will see separated oil and depending the density of the oil and refrigerant it will be on top or on the bottom.
We assume that the refigerant is heavier than the oil and thus oil starts floating on top of the refrigerant.
As this process continues, the liquid level in the heat exchanger rises because the density of the liquid is less in this leg of the communicating vessels.
If we assume that the density of the oil is 0.9kg/dm3 and that of the refrigerant is 1.3kg/dm3, than the difference in height will be like 45% of the height of the oil colomb on top of the refrigerant/oil mix.
So when there is 1m oil floating on top of the refrigerant/oil mix, than the level in H07.04 is 0.45m higher than in the separator.

So far so good, but now what?
Should the oil level reach the outlet of the heat exchanger at this point so that the oil flows to the compressor?
In that case the suction line should not be like on my drawing but at the same level as the liquid in the separator which I think is not a good idea.
How do I get that oil up say two meters up to the main suction line?
Should I install a check valve in the liquid supply, e.g. after V07.04 so that the pressure in the heat exchanger can rise to push the oil into the suction line (like with a coffee machine)?
That might work I think.

I could make a branch somewhere high in the heat exchanger and pipe that straight to the compressor suction with a solenoid valve to open when that compressor is running.

The level in the separator must always be the same for this system, e.g. HP float valve arrangements not allowed, am I correct?

It won't be easy to calculate the capacity of this recitifier.
It must be sensitive for variations in separator level
It must be sensitive for density:
>different refrigerants have digfferent densities (although syntetic refrigerants are mostly 1.3-1.4kg/dm3)
>some refrigerants like CO2 have very different densities at different emperatures
>different oil have different densities too, from AB oil (0.9kg/dm3) to PAG (1.05kg/dm3)

Hence, the design is good for one specific refrigerant oil combination at specific conditions?
Does anyone have a P&ID or mechanical drawing of a proven good working system?

Nico
05-01-2013, 12:32 PM
"The still which is maintained at a slightly higher pressure than suction"

Ranger, how does the liquid flow to the heat exchanger and to the still when the pressure in the still is higher, or do you use something like a gas pump with discharge pressure?

Correction:
Oeps, I see that the still is only for the oils separators in the discharge.

sandybapat
07-01-2013, 03:56 AM
Hi Nick

I could not find the Cl2 liquifier leaflet.

Sandybapat

Tycho
08-01-2013, 01:59 PM
The design with the connection to the pump and with the expansion valve that you refer to makes sence to me Tycho.
Proper control under various conditions, easy to calculate and posibility to distribute to several compressors.
I would do it that way too.:)

However, I like to know about the "troublesome" design where the density of the oil and refrigerant are crusial for proper working of the oil rectifier.
This is what I understood design wise so far (see attached drawing).

I have understood the idea is this (please correct me if I am wrong):
We look at a running system, at least one compressor is running, motor valve V04.02 is open for >15%, solenoid V07.02 is open.
Refrigerant with oil will flow into heat exchanger H07.04 and temperature of the refrigerant/oil mix starts rising.
Refrigerant boils off and that gas returns to the compressor through the piping on top.
The concentration of the refrigerant with oil goes up, until it reaches the maximum miscibility for that refrigerant with that oil at that temperature and that pressure.
If more refrigerant is boiled off than the we will see separated oil and depending the density of the oil and refrigerant it will be on top or on the bottom.
We assume that the refigerant is heavier than the oil and thus oil starts floating on top of the refrigerant.
As this process continues, the liquid level in the heat exchanger rises because the density of the liquid is less in this leg of the communicating vessels.
If we assume that the density of the oil is 0.9kg/dm3 and that of the refrigerant is 1.3kg/dm3, than the difference in height will be like 45% of the height of the oil colomb on top of the refrigerant/oil mix.
So when there is 1m oil floating on top of the refrigerant/oil mix, than the level in H07.04 is 0.45m higher than in the separator.

So far so good, but now what?
Should the oil level reach the outlet of the heat exchanger at this point so that the oil flows to the compressor?
In that case the suction line should not be like on my drawing but at the same level as the liquid in the separator which I think is not a good idea.
How do I get that oil up say two meters up to the main suction line?
Should I install a check valve in the liquid supply, e.g. after V07.04 so that the pressure in the heat exchanger can rise to push the oil into the suction line (like with a coffee machine)?
That might work I think.

I could make a branch somewhere high in the heat exchanger and pipe that straight to the compressor suction with a solenoid valve to open when that compressor is running.

The level in the separator must always be the same for this system, e.g. HP float valve arrangements not allowed, am I correct?

It won't be easy to calculate the capacity of this recitifier.
It must be sensitive for variations in separator level
It must be sensitive for density:
>different refrigerants have digfferent densities (although syntetic refrigerants are mostly 1.3-1.4kg/dm3)
>some refrigerants like CO2 have very different densities at different emperatures
>different oil have different densities too, from AB oil (0.9kg/dm3) to PAG (1.05kg/dm3)

Hence, the design is good for one specific refrigerant oil combination at specific conditions?
Does anyone have a P&ID or mechanical drawing of a proven good working system?

I am assuming R-22 as you mentioned in your first post, and as long as you have a stable system, with "normal" oil loss from the compressors, the highest concentration of oil wil be in the top layer of the *****, and the oil rich ***** will be lifted into the oil return pipe every time the motor valve opens and passes warm liquid through the exchanger, so the return you are getting is not pure oil, but oil rich *****.
The exchanger does not have a higher pressure than the suction, but the venturi effect or what it's called, where the oil return is connected to the main suction will help "suck" the oil rich ***** that is lifted up in the pipe.

The problem with the design in your draing is that the S05 drum has to have a very stable liquid level, and the exchanger needs a 60-70% level to be able to lift the oil, and the lifting height should be as short as possible.
The systems I have used it on, the lifting height has been 1.5 meters.
This design I have only seen on flooded chillers and would never work on a system where the LP drum has a variable liquid level, the one I showed you with the TX valve of course works on systems with variable level.
There is one more design that does not involve a TX valve wich can also be used on systems with variable level, but I will have to make a drawing of that later :)

One more problem with this design is if you are so unlucky as to have high oil carryover from the compressors, the exchanger may get swamped with oil.
Oil wil not, as you mentioned above, become a column on top of the ***** level, instead more and more oil will gather in the exchanger as the ***** is boiled off, and in the end you will have to shut it off and somehow drain it.

Nico
08-01-2013, 04:09 PM
Thanks Tycho very clear now.
This was a solution for single simple gravity flooded systems.
I assume that some sort of heating of that refrigerant/oil mix was standard (like I put in the discharge line).
This will not work when the separator has changing levels like when aircoolers are installed which need defrosting, or with batch processes like blast freezers.
In such case the level in the separator may rise very quickly and a nice bucket of refrigerant with oil from the other leg will end on top of your compressor pistons.

If you prefer to install the system with the expansion valve but there is no pump to force the liquid through the expansion valve like the case with gravity flooded systems, than there is still the option of using ejectors (bit complicated to control) or you can use a gas pump with discharge gas, that works like a charm.