Ammonia liquid carryover to screw compressor

[mariano-arg]

We have wrecked a compressor for liquid carryover. OEM says separator and suction piping sizing is perfect. Basically says operation or ammonia load was wrong.
Question is : how to check carryover calculation ? I mean checking the separator sizing and/or vapor speed in the suction line.
2nd question is : what is the right protection for the equipment in case OEM is right ? is suction superheat a good protection ? What should be the value ?

This is a York, chiller PAC 128 HR, rotatune compressor with frequency variator, cooling water from 14,5°C to 9°C, 490 MCal/hr capacity.

[Josip]

Hi, mariano-arg :)


We have wrecked a compressor for liquid carryover. OEM says separator and suction piping sizing is perfect. Basically says operation or ammonia load was wrong.
Question is : how to check carryover calculation ? I mean checking the separator sizing and/or vapor speed in the suction line.
2nd question is : what is the right protection for the equipment in case OEM is right ? is suction superheat a good protection ? What should be the value ?

This is a York, chiller PAC 128 HR, rotatune compressor with frequency variator, cooling water from 14,5°C to 9°C, 490 MCal/hr capacity.

yes, I've been there ... whatever is delivered by OEM always is perfect ....so you made mistake and you will pay for it ... otherwise they have to pay for mistake ... very simple

what about protection devices? which control you have on that unit UNISAB II or another?

it is possible that you charge too much ammonia, but in that case you have protection devices which will stop compressor

but, then if your control device setup was wrong ....?

Give us some more data, please.

Best regards, Josip :)

[mariano-arg]

thank you Josip,

Yes, control panel is Unisab II.
It does not have any obvious protection against liquid carryover. It does not have a level switch, for instance. It does not have any liquid sensor in the suction line, or anything else that could "boil" liquid between separator and compressor.
Suction pressure interlock is set up over 8 kg/cm2.
Suction superheat is set up at 0,5°C. Trips off the unit if set up over this value.
Discharge gas superheat is set up at 20°C w/5 minutes delay.
Other standard protections, but nothing obvious to protect from liquid carryover.
What could be a good protection ?

[RANGER1]

Does it have a suction trap in suction line ?
Normally a high level float to protect it entering compressor .
If its critacally charged suction trap should hold any excess refrigerant if TX valves fail .
Could be overcharged ?

[charlie n]

Suction superheat is set up at 0,5°C. Trips off the unit if set up over this value.
Discharge gas superheat is set up at 20°C w/5 minutes delay.
Other standard protections, but nothing obvious to protect from liquid carryover.


This is protection from liquid carryover. It sounds like the unit is overcharged or you have a very high return water temperature at start up. This can cause the ammonia in the evaporator to "burp" and slug the compressor.

[josei]

Mariano:

Carryover could be a cascade of problems.
Please look at my small contribution. PAC-chiller (http://j.mp/PAC_Chiller)

...

[Grizzly]

Mariano:

Carryover could be a cascade of problems.
Please look at my small contribution. PAC-chiller (http://j.mp/PAC-Chiller)

...
Interesting link there Josei.
I am on my way out of the door but will read more when I get back.
Thanks.
Grizzly

[piyush_ramani]

Can anyone have any idea about to design Low pressure receiver,or selection programme,

[mariano-arg]

Charlie n, low suction gas superheat should trip off the compressor, trying to catch liquid carryover. What is a good value for this protection ? half degree looks too close to the saturation curve.

Josei, interesting poing, by looking at the sight glass of the NH3 tank I can see some foam in the surface. Is this foam normal and expected ?

There is no trap in the suction line. There is no level switch in the suction line, nor in the NH3 tank.

This suction line is 4" starting horizontal and then down to the compressor. Compressor swept volume is up to about 1,500 Nm3/hr. What is the maximum velocity that this line should have to avoid droplets carryover?

[Josip]

Hi, mariano-arg :)


thank you Josip,

Yes, control panel is Unisab II.
It does not have any obvious protection against liquid carryover. It does not have a level switch, for instance. It does not have any liquid sensor in the suction line, or anything else that could "boil" liquid between separator and compressor.
Suction pressure interlock is set up over 8 kg/cm2.
Suction superheat is set up at 0,5°C. Trips off the unit if set up over this value.
Discharge gas superheat is set up at 20°C w/5 minutes delay.
Other standard protections, but nothing obvious to protect from liquid carryover.
What could be a good protection ?


What about alarms ... check measured values


Sabroe sometimes advice to put SH to 0K only for this kind of chiller with PHE ... strange but truth, in that case your suction superheat protection is deactivated .... otherwise your compressor will stop very often ..., but there is another better protection by discharge superheat....


There in UNISAB II version 2.05 is protection for liquid slugging warning and alarm ...

... warning is issued by a sudden decrease in discharge temperture which is greater then default 5C for a fixed 5 second period...

.... alarm is issued by a sudden decrease in discharge temperature which is greater then default 10C for a fixed 5 second period.

Then we have some other tools to regulate capacity, run time, timers for delay up or down, suction ramp etc... but that is not easy to explain here

Finally, first start up should be done by York supervisor and he must write down all set points i.e. values .... also take care about ammonia charging .... happen like that or not;)

UNISAB II is the most powerful PLC for regulation and control of refrigeration compressors, but set up must be OK.... otherwise is waste of money ;)

Best regards, Josip :)

[charlie n]

Mariano-arg, Josip has pretty well answered the question you had for me. The PAC chillers have flooded evaporators. The theoretical suction superheat is 0. Setting the suction superheat trip at 0.5 gives very little protection but if your superheat is higher, the chiller is undercharged or the PHE is fouled with oil. Discharge superheat is a better protection but neither will protect from an overcharged chiller. I installed and serviced several Sabroe PHE based chillers in the 90's before they had the pretty name (PAC) and before they got the oil return worked out. Some had no indication of liquid level so I would charge the unit with ammonia based on superheat. High superheat .. not enough charge. 1 degree superheat .. OK. For service ... high superheat .. oil in the Plate exchanger.

[Magoo]

slow down the capacity load rate, to stop boil up in accumulator and carry over to compressor.
magoo

[mariano-arg]

Mariano-arg, Josip has pretty well answered the question you had for me. The PAC chillers have flooded evaporators. The theoretical suction superheat is 0. Setting the suction superheat trip at 0.5 gives very little protection but if your superheat is higher, the chiller is undercharged or the PHE is fouled with oil. Discharge superheat is a better protection but neither will protect from an overcharged chiller. I installed and serviced several Sabroe PHE based chillers in the 90's before they had the pretty name (PAC) and before they got the oil return worked out. Some had no indication of liquid level so I would charge the unit with ammonia based on superheat. High superheat .. not enough charge. 1 degree superheat .. OK. For service ... high superheat .. oil in the Plate exchanger.

charlie n :
Ok, I accept it. Theoretical superheat is 0. I put it in manual mode and the unit runs 2 to 3°C superheat. Makes me comfortable. The minute I put it to run in AUTO mode, the capacity fluctuations makes superheat go to under 1°C. If I was running close to 0°C what would happen to superheat when capacity changes ??
It would hit the zero !! would that mean it is carrying droplets ??

[mariano-arg]

Hi, mariano-arg :)
There in UNISAB II version 2.05 is protection for liquid slugging warning and alarm ...

... warning is issued by a sudden decrease in discharge temperture which is greater then default 5C for a fixed 5 second period...

.... alarm is issued by a sudden decrease in discharge temperature which is greater then default 10C for a fixed 5 second period.

Then we have some other tools to regulate capacity, run time, timers for delay up or down, suction ramp etc... but that is not easy to explain here

.
.
.
Best regards, Josip :)

Josip, that kind of protection is exactly what I am looking for. I have Unisab II, don't know the version. How do I turn this function on ??

[Josip]

Hi, mariano-arg :)


Josip, that kind of protection is exactly what I am looking for. I have Unisab II, don't know the version. How do I turn this function on ??



Version you can read from UNISAB II and settings are described in UNISAB II manual ... you should have it!?!

Best regards, Josip :)

[charlie n]

charlie n :
Ok, I accept it. Theoretical superheat is 0. I put it in manual mode and the unit runs 2 to 3°C superheat. Makes me comfortable. The minute I put it to run in AUTO mode, the capacity fluctuations makes superheat go to under 1°C. If I was running close to 0°C what would happen to superheat when capacity changes ??
It would hit the zero !! would that mean it is carrying droplets ??

Mariano, I think you answered your own question. A lot of the problem is most likely coming from the load fluctuation. Take a look at this side of the system & try to find a way to buffer the load variations. Buffer tank or mixing valve or similar.
You can also try to slow down the load/unload rate of the compressor. As I remember Unisab, you can widen the P band and you could reduce the oil feed to the slide valve by closing the needle valve a little.
I wouldn't be comfortable with 3 degrees superheat. it sounds too high for a flooded evaporator.

[chilldis]

Sorry for my ignorance, but I have never encountered a ammonia system without a suction accumulator with a high level float switch or electronic level sensor. Is this design, without the safeties, common?

[josei]

Sorry for my ignorance, but I have never encountered a ammonia system without a suction accumulator with a high level float switch or electronic level sensor. Is this design, without the safeties, common?


Chilldis,

Look at bottom pictures. PAC chiller with suction acumulator.

As can see, HL switch not rqrd, bcause max. level ammonia (exagerated) is given by amount of liquid (critical) in system.
As any flooded system is always wet, not superheated.
This model produced by Sabroe-York-Frick, has sold thousands a.o.the.world.
Screw compressor + UNISAB, best to capacity control from minimun to 100% volumen.

:cool:

[chilldis]

Thank-you for the education Josei. That just proves that a person only stops learning when they are dead. thanks again.

[lowtempman]

I am a late comer to this board but all you guys have good thoughts and comments. I would only add that if the slopover is not due to rapid load increases causing "old faithful" to errupt then it is very likely that there is too much oil in the PHE. Note this is where all the oil that escapes the compressor oil separator will eventually end up. Do you keep track of how much oil you add to the system? Do you remove the oil from the PHE and record the amount? Does system have an automatic oil return system to get the oil back to the compressor that does not work? Years ago I had a trouble shoot job where the owner had added oil for 20 years to an old recip ammonia flooded shell and tube evaporator. He did not know he had to manually drian the oil and the evaporator was 1/2 full of oil. Approximately 300 gallons oil were removed from the flooded evap..