PDA

View Full Version : Cryo Refrigeration Help







Cryo99
21-07-2009, 10:47 PM
I am a process engineer new to a cryogenics unit. I am currently experiencing trouble with the refrigeration section of the unit. The cryo refrigerant system was installed in 1985 and originally used R-22 as the refrigerant. In 1997, the refrigerant was changed to refinery grade propylene (~70/30 propylene/propane) and subsequently changed to 99+% propylene. The system is driven by two (a booster and a high stage) Frick rotary compressors (RWB II - 316 and RWB II - 399). These compressors use a moveable slide valve and slide stop to control the load on the system. The booster compressor discharges to the high stage, which discharges into an oil separator/surge reservoir. From this drum, the refrigerant goes through a refrigerant condenser, which uses cooling water on the utility side. The refrigerant then goes through a "rock" filter/drier before discharging into a high pressure receiver, which essentially acts as a surge vessel. The refrigerant then exchanges heat with (i.e. is cooled by) the cryo recovered liquid leaving the unit before going to the economizer. As the refrigerant enters the economizer, its pressure is dropped across the economizer level controller. The economizer has a pressure controller that takes vapor back to the high stage suction. Liquid leaves the economizers and goes to the low pressure receiver where it again expanded across the low pressure receiver level controller. The liquid in the low pressure receiver goes to the quadraplex exchanger by thermosiphon. In the quadraplex the refrigerant cools the inlet process gas for the cryo and vaporizes back to the low pressure receiver. Vapor from the low pressure receiver is then sent back to the booster compressor to begin the refrigerant cycle again.

The problem I am seeing is that there doesn't appear to be room to run the booster compressor, which is limiting recovery in the cryo. When the booster is started the high stage hits a high discharge pressure limit and can't load up, which leads to a high discharge pressure on the booster machine. This high pressure then cause the booster to hit a high amp limit because it can't get enough load to "satisfy" the machine, which is observed as high vibration.

A similar problem occurred last summer and the high discharge limit was raised. After that change the system worked fine, but this summer it doesn't seem to be enough.

My thought is that the refrigerant condenser is fouled or under performing, so we aren't able to get adequate cooling to condense the additional refrigerant circulation when both machines are on. Still, I am wondering if there are any other likely problems that I might be overlooking.

Its probably a reach, but the one thing I am aware of that is different than normal is that the refrigerant (liquid) levels in the economizer and low pressure receiver are higher than usual (but still <50%). I don't suspect any liquid carryover to the compressors. Could too much refrigerant in the system lead to this problem?


(Sorry for the long post, but I appreciate any help or guidance.)

Gary
21-07-2009, 11:46 PM
My thought is that the refrigerant condenser is fouled or under performing, so we aren't able to get adequate cooling to condense the additional refrigerant circulation when both machines are on. Still, I am wondering if there are any other likely problems that I might be overlooking.

That condenser would be my primary suspect as well. I would check the water in and out temps, the saturated condensing temp, and the leaving refrigerant liquid line temp.


Its probably a reach, but the one thing I am aware of that is different than normal is that the refrigerant (liquid) levels in the economizer and low pressure receiver are higher than usual (but still <50%). I don't suspect any liquid carryover to the compressors. Could too much refrigerant in the system lead to this problem?

Has refrigerant been added to the system recently?

The danger is not so much liquid carryover to the compressors as it is excess liquid backing up into the condenser, thus limiting the condenser's ability to reject heat.

Cryo99
22-07-2009, 12:32 AM
Yes, due to some work we lost refrigerant levels. When we recharged the system I think we might have put too much (or at least more than we previously had).

There is a high pressure receiver between the economizer and the refrigerant condenser that does not have a liquid level, which is normal from the time I have been watching the unit. Since there this vessel isn't flooded, I don't believe that there is liquid backing into the condenser.

The reason I was curious about the levels is that I have observed high pressure (with just one compressor running) until level gets established in the LP receiver. I assume though that this is because, until there is a liquid level there, we are "short circuiting" the system and not getting the cooling from throttling down at the inlet to the LP receiver.

Gary
22-07-2009, 03:16 AM
When the booster is started the high stage hits a high discharge pressure limit...

Which points directly at the condenser. Non-condensables, overcharge, fouled tubes, insufficient water flow, high incoming water temp, etc.

The temp measurements I mentioned earlier will tell the story.

RANGER1
22-07-2009, 08:40 AM
To me the liquid reciever has to maintain a liquid level at all times , otherwise economizer will suffer as well .

ie flash gas and system inefficiency .

As Gary has mentioned non condensibles are of concern as well as condensor efficiency .
It could also be the rock filter drier banking up liquid in condensor . Is there a bypass or has anyone changed them + cleaned filter if it has one ?
Normally filters are on liquid line from liquid reciever to economizer .

Have had instances also if liquid reciever has loss of liquid seal that discharge pressure can go through the roof probably due to to much velocity for refrigerant passing through condensor to change state ( condense ) .

You may also have to investigate higher than normal levels in economizer and low temp accumulator as this is an area also that will be storing excess refrigerant that should be in liquid reciever .
Can you shut liquid makeup from reciever to economizer with stop valve and see what happens to levels and pressures .

What type of level control system is installed on economizer + low temp accumulator ?

Cryo99
22-07-2009, 09:14 PM
That condenser would be my primary suspect as well. I would check the water in and out temps, the saturated condensing temp, and the leaving refrigerant liquid line temp.

CW inlet T = 79F
CW outlet T = 87F
Refrigerant inlet T = 135F
Refrigerant inlet P = 217.5psig
Refrigerant outlet T = 94F
Refrigerant outlet P = 204psig

The outlet T/P falls on or just above the vapor pressure curve. I got results at a lower load and they were mvoed down the curve. I believe this means that it almost has to be a cooling limit rather than another problem. Is that correct?

The compressor discharge seems to have ~40F of superheat. Does that seem right for the screw compressor?

Cryo99
22-07-2009, 09:22 PM
To me the liquid reciever has to maintain a liquid level at all times , otherwise economizer will suffer as well .

Are you talking a level in the HP receiver? I am sure you are correct, but for the year I have been watching the unit it is rare to have a level here. The lines to and from this vessel are on top, so should this vessel be flooded? If not how does the liquid leave.



As Gary has mentioned non condensibles are of concern as well as condensor efficiency .

While I am not positive, from the data in the previous post, I think we are on the curve, so I don't expect non-condensables. If they are a problem it is only in the condenser as we took a sample off a high part of the unit and things looked normal (99% propylene).


It could also be the rock filter drier banking up liquid in condensor . Is there a bypass or has anyone changed them + cleaned filter if it has one ?
Normally filters are on liquid line from liquid reciever to economizer .

The rock fitlers were changed last week and we didn't see any improvement. There doesn't seem to be any unexplained DP across the rock filters.


You may also have to investigate higher than normal levels in economizer and low temp accumulator as this is an area also that will be storing excess refrigerant that should be in liquid reciever .
Can you shut liquid makeup from reciever to economizer with stop valve and see what happens to levels and pressures .

What type of level control system is installed on economizer + low temp accumulator ?

I lowered the levels in these vessels and didn't really see any change other than the compressor unloading. I also didn't gain level in the HP receiver, so I am not sure where the refrigerant went, but...

Gary
22-07-2009, 09:41 PM
CW inlet T = 79F
CW outlet T = 87F
Refrigerant inlet T = 135F
Refrigerant inlet P = 217.5psig
Refrigerant outlet T = 94F
Refrigerant outlet P = 204psig


I don't have a pressure/temperature chart for propylene, which is why I requested saturation temperatures rather than pressures.

Cryo99
22-07-2009, 10:42 PM
Sorry. The condenser outlet is on the curve (or just on the liquid side) depending on the accuracy of the measurements.

At the condenser outlet P, the sat T is 96F.

Gary
22-07-2009, 11:30 PM
CW inlet T = 79F
CW outlet T = 87F
Refrigerant inlet T = 135F
Refrigerant inlet P = 217.5psig
Refrigerant outlet T = 94F
Refrigerant outlet P = 204psig/96F


87-79 = 8F dT. Water flow is adequate.

96-79 = 17F TD. Load not excessive... if anything light load.

96-87 = 9F approach. Condenser not fouled. Non-condensables unlikely.

96-94 = 2F SC. Not overcharged... if anything undercharged.

I find the pressure drop across the condenser (217.5-204 = 13.5psi) puzzling.

And we still have no explanation for the system kicking out on high discharge pressure. What is the control's pressure/saturation temp setpoint?

RANGER1
22-07-2009, 11:39 PM
Do you have a P&ID of plant to show us your system ?

High pressure Liquid recievers can have a dip pipe down to bottom of vessel for liquid take off to economizer , maybe you have this ? Rerigerant drains into HP reciever then liquid is fed from i presume other pipe on top of HP reciever to economizer and intercooler if you have one

Non condensibles are usually found in condensor liquid drain when plant is running and at top off condensor while off and cool .
You can purge non condensibles from liquid drain while running to see if any change ( if required ) . Maybe worth a try even if temps / pressures don't show signs of non condensibles .
I would have thought you would also have a intercooler of some kind between booster discharge and second stage compressor suction ?

Cryo99
23-07-2009, 12:31 AM
87-79 = 8F dT. Water flow is adequate.

96-79 = 17F TD. Load not excessive... if anything light load.

96-87 = 9F approach. Condenser not fouled. Non-condensables unlikely.

96-94 = 2F SC. Not overcharged... if anything undercharged.

I find the pressure drop across the condenser (217.5-204 = 13.5psi) puzzling.

And we still have no explanation for the system kicking out on high discharge pressure. What is the control's pressure/saturation temp setpoint?

All of this data was with only the high stage compressor running (and fully loaded). As I loaded the compressor up, the discharge pressure and temps came up, but everything moved up along the vapor pressure curve. (I can post the less loaded data if helpful.)

I can also try to get some data with the booster running (but it might not be steady state as I am not sure we can keep it running that long). The data with the booster might have to wait a week or so though, as I will be out of the plant for a while.

Cryo99
23-07-2009, 01:53 AM
Do you have a P&ID of plant to show us your system ?

High pressure Liquid recievers can have a dip pipe down to bottom of vessel for liquid take off to economizer , maybe you have this ? Rerigerant drains into HP reciever then liquid is fed from i presume other pipe on top of HP reciever to economizer and intercooler if you have one

Non condensibles are usually found in condensor liquid drain when plant is running and at top off condensor while off and cool .
You can purge non condensibles from liquid drain while running to see if any change ( if required ) . Maybe worth a try even if temps / pressures don't show signs of non condensibles .
I would have thought you would also have a intercooler of some kind between booster discharge and second stage compressor suction ?

I will try to put together a block flow if that will help. I can't post our P&IDs. Until then, he refrigerant goes through the condenser (in the top out the bottom), runs at ground level through the rock filter, and up to the top of the HP liquid receiver. It leaves through another line from the HP goes through a sub cooler, which is heat exchange with the recovered cryo liquid product. Again the flow is in the top of the exchanger and out the bottom (the exchangers are horizontal, of course). The outlet flow then goes through the economizer level control valve and into the economizer.

I don't believe that we have the dip leg in the HP receiver, but I will check. That would explain a lot as I have never understood what the HP receiver really was for, since there is rarely a level.

The sample we took (while running) was vapor off of the HP receiver, so if that is where the non-condensibles would collect, I don't think that is the problem, but we could vent to be certain.

There is no cooler between the two compressors. The discharge of the booster mixes with the economizer gases and goes directly into the suction of the high stage compressor. I believe I have read that screw compressors are better about the amount of heat of compression that goes to the gas (as much of it is absorbed by the lube oil), which is why I think the interstage cooler is not required.

RANGER1
23-07-2009, 03:52 AM
Do you have a point to purge non condensables directly after condensor outlet as this is where it will collect as it will be trapped before "P" trap

RANGER1
23-07-2009, 09:50 AM
High pressure liquid reciever is for several reasons to clarify .

1/ To store excess refrigerant from plant if a pump down is required .

2/ To always have enough refrigerant to satisfy all levels in vessels like economizer and LP reciever under varying conditions / loads .
Usually 25% or so minimum level .

Question
what type of condensor do you have , ie shell + tube , plate heat exchanger ?

Cryo99
23-07-2009, 01:20 PM
The HP receiver does have the stand pipe that you described, which makes the operation of the vessel make more sense.

The condenser is a shell and tube exchanger. There is not a place to purge the outlet until it gets to the HP receiver.

Cryo99
23-07-2009, 04:11 PM
We had some cooler weather this morning and managed to get the booster online (for now).

With just the high stage running (fully loaded) we had

CW inlet T = 75F
CW outlet T = 81F
Refrigerant inlet T = 133F
Refrigerant inlet P = 204psig
Refrigerant outlet T = 89F
Refrigerant outlet Sat T = 94F
Refrigerant outlet P = 193psig

We then unload the high stage some and started the booster. With everything loaded back up as much as we could the numbers were

CW inlet T = 76F
CW outlet T = 86F
Refrigerant inlet T = 157F
Refrigerant inlet P = 237psig
Refrigerant outlet T = 98F
Refrigerant outlet Sat T = 102F
Refrigerant outlet P = 222psig

This is with both compressors running and satisfied. When the ambient temp goes up today and we run into our "normal" problem, I will take another set of measurements.

I did notice that since we moved the levels around and loaded the high stage fully, the level in the economizer is very low and still no level in the HP receiver. This is more what I am used to seeing, but based on RANGER1's comments, I am wondering if still more Ps wouldn't help.

Gary
23-07-2009, 04:20 PM
What is the setting for the high pressure cut-out?

Cryo99
23-07-2009, 04:40 PM
The high stage machine will stop loading at 235psig, unload at 240psig, and shutdown at 250psig. As you can tell we are pushing this already.

Gary
23-07-2009, 04:41 PM
...still no level in the HP receiver. This is more what I am used to seeing, but based on RANGER1's comments, I am wondering if still more Ps wouldn't help.

102 - 98 = 4F SC. This should be 10-15F, so clearly more refrigerant should be added and this would give us the proper level in the receiver.

But we still need to find out why it is going out on high pressure. It could be as simple as the setting not high enough... or there could be a problem we are not seeing yet.

Gary
23-07-2009, 04:42 PM
The high stage machine will stop loading at 235psig, unload at 240psig, and shutdown at 250psig. As you can tell we are pushing this already.

What saturation temps do these pressures correspond to?

Cryo99
23-07-2009, 04:44 PM
I did notice that since we moved the levels around and loaded the high stage fully, the level in the economizer is very low and still no level in the HP receiver. This is more what I am used to seeing, but based on RANGER1's comments, I am wondering if still more Ps wouldn't help.


Of course, the levels I am seeing now are what I am used to seeing when both machines are running well. That doesn't mean they are right, but they are normal.

Cryo99
23-07-2009, 04:53 PM
What saturation temps do these pressures correspond to?


235psig - 108F
240psig - 110F
250psig - 113F

Cryo99
23-07-2009, 04:56 PM
235psig - 108F
240psig - 110F
250psig - 113F

But, there is about a 15psi pressure drop across the condenser, so that lowers the temperature about 5F (i.e. a compressor discharge of 235psig would give a condenser discharge of ~220psig or 103F sat temp).

Cryo99
23-07-2009, 05:01 PM
We are staying on the saturation curve, but moving up and down it with load changes. Yet we aren't building level in the HP receiver or economizer. Does that not mean that we are vaporizing it in the LP receiver as fast as we can condense it?

Cryo99
23-07-2009, 05:02 PM
We are staying on the saturation curve, but moving up and down it with load changes. Yet we aren't building level in the HP receiver or economizer. Does that not mean that we are vaporizing it in the LP receiver as fast as we can condense it?

Or is it a sign that we are undercharged?

Cryo99
23-07-2009, 05:03 PM
Here is a block flow of the system.

http://img216.imageshack.us/img216/824/refskid.th.png (http://img216.imageshack.us/i/refskid.png/)

Gary
23-07-2009, 05:11 PM
235psig - 108F
240psig - 110F
250psig - 113F

These settings are not high enough for a warm day, much less a hot day. I would expect a high pressure cut-out to be set up in the neighborhood of 130-135F or even a little higher.

And the pressure drop across the condenser is still puzzling. Are there any other devices between the two pressure measurement points?

Cryo99
23-07-2009, 05:44 PM
These settings are not high enough for a warm day, much less a hot day. I would expect a high pressure cut-out to be set up in the neighborhood of 130-135F or even a little higher.

And the pressure drop across the condenser is still puzzling. Are there any other devices between the two pressure measurement points?

Unfortunately, 130F corresponds to a pressure of 305psig, which is above our system RV setpoint. My guess this is the downside of the refrigerant change they did many years ago.

Still, the system worked last year with these same settings and during days of hotter weather than this. Could the explaination be that the condenser performance isn't horrible, but not as good as it was last year?

Other than the oil seperator there isn't anything in the line between the two points besides pipe and the condenser. The gauge on the outlet of the oil sep is about 2-3psi lower than the compressor discharge P.

Gary
23-07-2009, 07:45 PM
Unfortunately, 130F corresponds to a pressure of 305psig, which is above our system RV setpoint. My guess this is the downside of the refrigerant change they did many years ago.

Still, the system worked last year with these same settings and during days of hotter weather than this. Could the explaination be that the condenser performance isn't horrible, but not as good as it was last year?

Other than the oil seperator there isn't anything in the line between the two points besides pipe and the condenser. The gauge on the outlet of the oil sep is about 2-3psi lower than the compressor discharge P.

I'm thinking there may be a restriction in the oil separator, possibly a plugged inlet screen if it has one... and if that restriction appeared recently that would explain why it worked better last year.

I'm guessing the pressure control is sensing the higher of the two pressures?

Is the RV on the condenser? If so, we could see a big improvement by moving the pressure sensing connection to the condenser out port.

But that does not resolve the restriction issue. You might want to contact the oil separator manufacturer to find out what their normal pressure drop should be.

Gary
23-07-2009, 07:51 PM
The gauge on the outlet of the oil sep is about 2-3psi lower than the compressor discharge P.

Hmmmm... then the drop is in the condenser. Maybe I'm wrong on this, but it seems like an excessive pressure drop for a condenser.

RANGER1
23-07-2009, 10:20 PM
It looks like propylene has similar pressures as R502 so if it was converted from R22 this could be a problem , especially with relief valve settings .
If the 2-3 psi pressure drop across oil seperator is same no matter what flow goes through it it could be non return valves coalescer elements .

If you havn't already , can you go through system and tell where you are taking the pressure reading exactly ?
You could consider getting refrigerant charge right .

Finding a way to purge non condensibles from condensor outlet to eliminate it ( maybe dump whats in whole condenser ) .

Sometimes shell and tube condenser water side divider plates can corrode also short cicuiting the water flow .
Has it been inspected since last year ?

Magoo
24-07-2009, 02:42 AM
If booster is unloading due to amps, it because the intercooler cannot handle the loads. If the high stage cannot maintain interstage design, then possibly non-condensibles or condenser duty/ capacity.
Changing refrigerants may be part of the problem. Current draw and pressures etc..,
Two stage systems are a fine line between operational and not.

Magoo
24-07-2009, 02:51 AM
Cryo99, can you breifly describe the start sequence, of high stage and then booster, you may get around issues by slowing load rates with booster to allow high stage to handle the load changes in interstage vessel. May extend pull down rate but will balance loads between stages.
magoo

RANGER1
24-07-2009, 04:15 AM
Magoo ,
There is no intercooler its booster discharge straight into HS comp suction ( see block diagram above ) .
We are trying to see whats different from last years performance i guess .

Looks like propylene will require A bit more power to do same job approx 14% from mycom programme .
Need an engineer to work out what happens to condensor performance .

Gary
24-07-2009, 06:24 AM
It seems pretty straightforward to me: Last year the system had a discharge restriction, causing the discharge pressure to rise and the system to cut off on high pressure.

They raised the pressure setting and it worked fine for the rest of the season.

This year it is again cutting off on high pressure, despite the higher setting.

What changed? The discharge restriction got worse.

Gary
24-07-2009, 06:40 AM
The condenser itself is a big barrel. It is not going to restrict anything.

If the restriction were at the condenser outlet it would act as a metering device and the leaving liquid would be much colder.

Apparently there is a relatively small pressure drop across the separator, so that rules out a restricted separator.

All that's left is the piping, fittings, valves, etc. between the separator and the condenser. Somewhere in there is the restriction. If there is a hand shutoff valve in this line, it would be the prime suspect.

RANGER1
24-07-2009, 08:56 AM
Thats why it would also be nice to know where exactly each pressure reading was also .
What type of reading accuracy as well .

Magoo
25-07-2009, 01:59 AM
Hi RANGER1,
thanks for correction, the block drg doesnt show oil cooler, filter or pump either, nor discharge stop/check. Any could create the base problem. Coverted from 22 originally, there would have been some sort of intercooling either vessel or liquid injection somewhere.
I am thinking that stop/check after high stage is partially restricting, Frick have a combo type vav, and can wear slide shaft and only partially open

RANGER1
25-07-2009, 02:37 AM
Agree with what you say except on intercooling .
For example a mycom compound uses only oil to cool interstage as well as some Grasso's pump directly from booster staight into high side suction .
On this plant he also has a side load from economizer suction .

If we know where pressure drop is when Cryo99 get back it may narrow it down as well .
He says there is no purge point between condenser outlet and liquid reciever , so he might be taking reading at HP reciever ?

Magoo
26-07-2009, 01:25 AM
Hi RANGER1,
an interesting problem, I have seen those Mycom compounded screws, but never really worked on them.
Oil temp is critical, Frick use Amot three port control vavs., and I have had a few of them play-up with age, which in turn put a larger load on condenser. It will be interesting where the HP pressure is being read from, the set or the condenser, I am picking it is from the set, from where all control functions are from.
As you said earlier the current refrigerant will put extra load on condenser as well, add all and any other problems will ramp HP problem
Come back to us all Cryo99.

Cryo99
03-08-2009, 07:35 PM
Okay, I am back.

The discharge pressure is measured at the compressor discharge before the oil sep or any valves by pressure transducer. Today (with only the high stage running) this pressure was 212psig. This is the pressure that the compressor shutdown is tied to.

The refrigerant then discharges into the oil sep, which contains coalescing elements, goes through a 3" check valve, a 3 90 angle valve, and a short section of 3 line before swaging up to a 6" line with a 6 manual valve. After this manual valve there is a pressure gauge, which today read 205psig. (Before I had said there was only a couple of psi drop between these two points. Apparently I read it wrong before. I am certain that today it showed a higher DP. Sorry for the misinformation.)

The 6" line then goes to the condenser with no valves in the line. The condenser outlet (a 3" line) then run to another pressure gauge before the rock filters. There are no valves in this line before the pressure gauge. Today this pressure was reading 203psig.

After the rock filter and the 3" manual valve on its inlet and outlet, the 3" line discharges into the HP receiver. Its pressure gauge read 198psig today.

In summary,

From compressor discharge to after all valves there is a 7psi drop.

Across the condenser there is a 3psi drop.

Across the rock filter and its isolation valves there is a 5psi drop.

Of course all of this is with different pressure gauges, so the DPs might be off. Assuming they aren't, is the 7psi drop in the oil sep and valves the most suspect? What is the best way to determine if the coalescing elements are plugged or if the check or 90 valve are malfunctioning?

Gary
03-08-2009, 08:25 PM
Is this a temprite oil filter/separator?

Seems like I recall that the temprites pop an o-ring and bubble violently in their lower sightglass (or was it the upper sightglass?) when the filters get plugged. I could be wrong on this. It has been a while.

Cryo99
03-08-2009, 08:38 PM
I am sorry. I don't know enough to answer for certain, but I don't believe it to be.

It is a system provided be Fricks and has 4 filter elements at the discharge end. Nothing in the BOM specifies Temprite, but it doesn't give more info than generic description of filter element along with a part number. There is not a sight glas at this end of the vessel (only bullseye to determine if oil is present). I have never noticed any bubbling in the sight glass at the opposite end.

RANGER1
03-08-2009, 09:39 PM
Unless there are points to measure pressure pressure drop before discharge check and stop valve it is near impossible to prove .
Can only suggest instal new coalesers as you usually can't see if theres any thing wrong as well as at same time disassemble check valve and stop valve .
If coalescers havn't been changed for awhile now would be a good time and log results if any .

Also rock filter looks suspicios as well if it has low rates . There should be a screen strainer in the assy as well which may need cleaning .

Cryo99
03-08-2009, 10:24 PM
The rock filter is just a porous "clay" media that is loaded into the housing. There is no additional strainer. There are two in paraller. Both were changed 2 weeks ago.