SullAir/Jordan Valve

[jaysephus]

personal stuff sorry

[NH3LVR]

I do not have much experience with Sullair compressors. Years ago they did not have a good reputation. In 1982 I ran three of them one summer. But since two of these machines are still running, they may have been better than people thought.
I saw no reference to a TX valve on Jordan's site. Of course that valve may have been installed long ago.

Onto the problem.
I would start by seeing what the superheat is at on the suction line. If it is falling rapidly that may indicate a slug. However I would think you would hear that
The valve may be opening too rapidly, but usually that would not lower the oil temperature that quickly.
I think you might want to put a thermometer on the suction line that is capable of recording a minimum temperature. That could eliminate the slugging theory.
What suction pressure do you operate at?
Send us a picture of the valve.

[RANGER1]

Jordan valves are from steam control valve type application .
Have seen the used before , they are equivilent to Danfoss TEAT control valve .
Check to see if it controls steady oil tempreture when not playing up .
Suggest it should control about 50-55 deg C discharge / oil temp .

If plant was a bit short of liquid for awhile it may open to far to try to control temp then when level ok it floods it
Seat in valve could be worn or damaged , you would have to pull it out + check .

Quantum panel should have some data logging to track what is happening .
If not you can get a data logger (one brand name is tiny tag ) and place it say on compressor rotor housing to see if liquid is coming down suction line into rotors .

[jaysephus]

personal stuff sorry

[jaysephus]

personal stuff sorry

[RANGER1]

I'm curious about ice making its way down rotor housing at 33 psi -19 f .

Are other machines upstream icing up the same ?

Gas may be a little wet for this to happen .

Could liquid pool at end of suction header where take off is to comp ?

As liquid builds up over time it might get liittle slugs as you describe .

Is there a drain valve off bottom of header to see if liquid there or insulation that can be removed ?

Pipe would be sticky if you put a wet finger on it ( don't get it stuck ) .

If you put data logger on instal it closer to discharge or where its has a bit of warmth , so any temp drop will be obvious .

Are all these compressors are pulling off same vessel , could level be suspect in vessel ?

Can you lower level in vessel ?

Have seen very similar problem where vessel was a bit small and it carried droplets over into header , it built up and slugged over every now and then .

Also depending on what type of liquid makeup you have , it may be open to far if manual expansion valve blasting liquid everywhere .

[Andy]

Hello Jaysepus:)

Maybe the liquid injection is the problem:confused:

Water or thermo oil cooling would be better. Is it possible to change to water cooled oil cooling?

Can the oriface size be reduced in the injection valve to reduce the possiblity of hunting?

Can the injection valve element be changed to control on the slightly lower oil temp, or can the bulb just be moved to the oil supply line, this will allow the valve to react to the low oil temps more quickly.

Kind Regards Andy:)

[jaysephus]

personal stuff sorry

[NH3LVR]

Ice forming at 19F on the compressors takes a low superheat. Dropping 20F in five seconds take a very wet return. I would want to check the suction temp sensor for the Quantum panel. Switch it out with a functioning one if you have a identical panel.
Another problem I have seen is if you have a malfunctioning power supply. Check the output voltage (DC) of the power supply and then turn your meter to AC. This will give you the AC ripple in the DC output. It should be below one or two percent if memory serves.

[jaysephus]

personal stuff sorry

[RANGER1]

Personally I wouldn't worry about heat trace on suction , as ice will still probably form anyway .
Normally ice will form there with little s/heat and it can get at a guess 1" thick .
Where is liquid injection entering , is it into rotor or discharge housing ?

I've been told liquid injection is about a 15% loss of efficiency , so depending on how big electric moter is power bill could add up pretty quickly if it is on 1 or more machines .

[jaysephus]

personal stuff sorry

[NH3LVR]

I think the latent heat of NH3 is way too much to evaporate a slug with heat tape.
These are the ones I used to enjoy going out on.

[jaysephus]

personal stuff sorry

[jaysephus]

personal stuff sorry

[Frickman]

about 150 F is the standard on a 95 F condensing temp. Frick acutaaly likes to see about 160 discharge temp, but with liquid injection they are normally set at 150.

[RANGER1]

jaysephus ,

I think 135 F (56 deg C ) sounds good and 115 F (46 deg C ) low ,but not the end of the world .
Usually work on saturated discharge of 95 F ( 35 deg C ) + oil temp to be higher 60 F ( 15 deg C ) so oil is not diluted with refrigerant .
Depending on oil type 135 F would be about maximum for mineral oil .

[Gabriel Minor]

In response to the Jordan valve question they do act just like a regular power head on a txv. What I do to test the power head is remove the sensing bulb from the discharge line and place it in a bucket of hot water. You should be able to see the linkage move to open the valve. The water does not have to be very hot about 140 degrees F. You should also do this while the compressor is off line. The other thing to look at is the old liquid injection solenoid valve before the Jordan valve. On the original systems there was a solid state temp. control that would not energize the solenoid valve. I would think that if either one of those components got stuck open (or manualy left open) that would cause the low discharge temperature issue that you are having. good luck

[US Iceman]

As far as I'm concerned Sullair built good refrigeration compressors. The Jordan valves inject liquid into a side port on the compressor prior to the rotors reaching discharge pressure. They can't inject into the discharge since the discharge pressure there would be slightly higher than the liquid injection pressure, right?

Ice on the compressors has to be put into context; that is , how low is the suction pressure? If the saturation temperature is below 32°F or 0°C, then it can form frost, which can eventually turn into ice. Just because a compressor has frost or ice on it does not mean liquid is present. You have to look at the discharge superheat (discharge temperature - saturation temperature at discharge pressure). IIRC Sullair used to run about 115-125°F oil injection temperatures. Therefore if you were running approx. 180 psig discharge (which equals about 95°F) you a discharge superheat of 125-95=30° to 115-95=20°. That is sufficient superheat to indicate you have no liquid in the compressor.

What usually happens during a slug is the discharge temperature will decreases very rapidly. When this occurs the discharge superheat will also drop quickly.

What you are describing sounds like an operational problem or piping issue. Let me ask one question: When the slugging occurs (whenever) is the compressor trying to load up?

[Segei]

You have 9 high stage compressors. How many compressors do you run? Shut this compressor off. If you have the slug, sooner or later other compressor should take it. If other compressor don't take slug, problem is Jordan valve. What does happen with mentioned compressor when it shut off? Is it load or unload? What is the capacity of this compressor?

[US Iceman]

Let me back up and ask a basic question. How are you defining a slug?

[jaysephus]

personal stuff sorry

[jaysephus]

personal stuff sorry

[Segei]

For me, it looks like Jordan valve is hunting. Probably it is oversized. Try to throttle valve before or after Jordan valve(better after). Compressor should be loaded 100%. Discharge and oil temperatures should be stable not jump up and down.

[US Iceman]

It's been a long time since I messed with one of those valves. They are actually a sliding gate valve. One is fixed and the power element moves the other IIRC.

What I would like to see for data is:

Oil temp / Disc Temp / disc press / Suction press / Suction temp

when one of the slugs occur.

If the Quantum panel picks up any deviation it's just spitting out a warning that some liquid is present in the compressor.

What we don't know if the liquid is coming from the Jordan valve or the compressor suction.

Has anyone worked on the Jordan valve before this started to occur?

Once we know if the liquid is coming from the liquid injection or compressor suction then we can track down what the cause is.

[Segei]

As I mentioned early, to test for slug this compressor should be shut off. Remaining recip. compressors are more sensitive to the slug than screw compressors. If you have slug, you will see it on recip. compressors as well.

[jaysephus]

personal stuff sorry

[RANGER1]

jaysephus ,

" Normally " oil and discharge temp are very close .
You have a control panel which is not original , giving readings that originally wern't monitored .
If it wasn't monitored no one would probably know or suggest anything was wrong .

Jordan valve may very well be starved every now and then giving erratic readings .

Newer plants these days usually have dedicated line with preferential liquid feed line as to avoid problems you are experiencing .

Have seen some similar erratic temps on discharge temp when compressor is starving for oil injection , so mixing of oil and refrigerant happens in oil seperator instead of discharge line .

It would also be intersting to monitor oil seperator oil temp to see if it fluctuates as much as discharge temps .

Insulating liquid injection line downstream of jordan will probably only stop water on floor ,but not much else ie flash gas or anything .

Rapid fluctuating head pressure could also erratic liq feed too .

[US Iceman]

Right, maybe I need to throttle it down. There is a globe isolation valve after the Jordan valve for maintenance.


For a short time, I think you could do this, but not for a long time. The flash gas will erode the valve seat if the globe valve is throttled for too long.
[/quote]



US Iceman I can tell you suction pressure is always 33 psi and it only varies plus or minus 2 but is usually pretty tight. Suction temp is something we don't log, but I did look at the couple freeze screen saves from the liquid slug alarms and they are all exactly the same, nothing showing a rapid drop.


The temperature drop you should expect to see o the suction line at the compressor is the difference between the actual gas temperature and the saturation temperature for 33 psig, if that's what you are running. You won't see a huge temperature reduction because the liquid is boiling at 33 psig, not something else (if the suction pressure is consistent).


The valve finds a happy medium and stays close the entire time, not fully open then closed and repeating but finding like a 61.23 percent open and kind of hangs out right there then very slightly going either way.


If the valve stem doesn't move a great deal that is what you want. That behavior indicates a stable valve. That's good!



If we are starving this valve and the discharge temp rises and the Jordan opens, and opens and opens trying to cool down the discharge. Then when it finally gets it's liquid the valve is too far open from what it really should be, and takes a quick shot of liquid and "slug alarm."


A distinct possibility. If this is occurring the discharge temperature would steadily increase or be elevated above normal. Then when a sufficient volume of liquid hits the Jordan valve it suddenly overfeeds because the valve is stroked fully open.

If the demand for liquid is high (check what is happening on the other users calling for liquid) and the Sullair is tail-end charlie on the liquid line you could have liquid distribution issues in the main liquid feed line to all of this.

I take it no other compressor has this issue?

[jaysephus]

personal stuff sorry

[RANGER1]

jaysephus , I would suggest manual expansion valves should be open 70 % of the time under fuul load conditions .


They can be adjusted seasonally and from your description of your plant it may be relevant .

Sorry didn't mean to put you off solving problem , sometimes you have to be there to see whats going on .

[jaysephus]

personal stuff sorry

[US Iceman]

The other things wanting to share this liquid is the cold storage intercooler and the cold storage recirculator vessel. So the other thing I can watch for is when the vessels are calling for liquid...


Not entirely... read below.



Would it make sense to insulate this high pressure liquid line? I don't think so but the SullAir manual says that to do that in cause of erratic temps. It's high pressure warm liquid so I don't see that working.


The Sullair manual IS correct. Think about it like this... If the liquid line is cooler than the ambient air temperature, which way does heat flow? It flows into the cooler liquid line. If the liquid line comes straight from the receiver it does not have any subcooling therefore the liquid can begin to flash off. The first compressor gets all of the liquid it needs, while the subsequent compressors only have access to the remaining liquid. As the liquid continues in length the remaining compressors get less and less liquid.

Then if for some reason the oil cooling load on one compressor decreases there is more liquid available for the remaining compressors downstream. If the Jordan valve is stroked all the way open when that occurs it has to throttle back. If it can't do this fast enough a sufficient volume of liquid may get into the compressor causing the alarms.

The flash gas in the liquid line is also caused by changes in pipe elevation. When a high pressure liquid line is installed to run vertically upwards the change in elevation reduces the liquid pressure due to static head penalties. This can be compounded installing a liquid line that is too small in diameter. This causes an increase in friction loss due to flow.

So... it is necessary to insulate a properly sized high pressure liquid line.

If you ever hear someone saying a 3/4" liquid line is sufficient for any reason you should ask them: based on what? And see what their answer is.

There are reason for why things work. Unfortunately, these are also the same reasons why things don't work.;)

[TXiceman]

My experience with the Jordan valve for liquid injections control is that it s a very slow acting valve. Bsing slow it will either overfeed or underfeed when you get a sudden oil cooling load.

I do not like the Jordan as a LI control. Have you considered changing to a Electronic control valve?

Ken

[RANGER1]

Another good point , Danfoss TEAT are mechanical and very good as well as ICM which is electronic modulating valve .
I'd say Hansen have equivilent .

Currently there are a lot of basic things to check like

-Balancing manual expansion valves on vessel makeups .

-Possibly insulating liquid line.
Have never seen it in my experience but also never encountered this problem .

-Checking for liquid flood back using logging device .


jaysephus , what type of liquid injection valves do you have on other compressors that seem to work OK ?

[jaysephus]

personal stuff sorry

[US Iceman]

Insulate the liquid line, sorry, if it were up to me it would, boss will not let that one fly.


Why? Is it money or doesn't he buy the fact that high pressure liquid can flash off?

One minor posting note... Can you add some spaces between you paragraphs please? I have a heck of a time trying to read everything all together like that.;)

Thank you....

[jaysephus]

personal stuff sorry

[US Iceman]

It doesn't help when he asks us, "well why does the sullair do that" and the dumb*#&@ that I work with say it's because the sullair is painted green, and that green paint, as we all know, is an anti-refrigeration color and thats why they dont make them anymore, or because the earth's gravitaional pull is different today because a cougar died today at a native american burial ground, or........


Kind of difficult to argue with that because it makes absolutely no sense. But I understand what you're facing. It's the same type of logic used by some who don't understand.

The only way you can prove the flashing is to install a sight glass like you mentioned earlier. For those that don't know the theory you have to show...

[Segei]

Measurement of pipe temperature will show flashing.

[RANGER1]

Armaflex type insulation for is resonably cheap and do it yourself .
You could buy small amounts at a time so no none gets upset .

[jaysephus]

personal stuff sorry

[Segei]

Armaflex is banned in our company, not kidding. Around 2001 someone put Aramflex on the cold storage intercooler, running at 30 psi, with no vapor barrier it ruined the vessel in 2 years and had to be replaced.

However, I will look into other low price insulation I can do myself.
I don't think that insulation will help you. Temperature difference between condensing(liquid ammonia)temperature and compressor room temperature isn't significant. Heat transfer coefficient from air(natural convection) to liquid line isn't significant either. Very little heat(if any) will be transfered to the liquid ammonia. Why to put insulation?
Did you determine the reason of mentioned problem with your compressor?
I think that problem with Jordan valve.
May be it is slow to react. To match this, compressor should load and unload slowly. Try this.
May be it is oversized. Throttle outlet valve. Did you try this action? It doesn't require significant investment.