NH3 thermosiphon oil cooler

Dear all,
We designed a NH3 thermosiphon oil cooler, shell and tube heat exchanger, ammonia boiling in tubeside, duty:300KW, tube counts:717, circulation ratio:3:1, oil inlet/outlet temperature:80/40 Deg C, ammonia: 38/39 Deg C.
But after running some days, we found that it didn't work well, the ammonia liquid has 2 Deg C sub-cooling, we measured some parameters as follows: duty:251KW, oil:87.7/49, ammonia:33.3/35.4, and if the entironment temperature becomes higher ,the oil cooler performance will become much worse.
We think the area we designed is less than the aera required.
could anyone help me to figure out this problem?? and give some advice on thermosiphon oil cooler designing?

thank you very much!


cygbob

Cygbob
A few questions.
Did it operate satisfactorily at first?
Is the Oil Cooler Horizontal or Vertical?
Pardon me if I misunderstood but are there 717 Tubes in the heat exchanger?
Do you have a sketch of the Thermosyphon System you could post?
Is there a lot of oil circulating in the plant?
Nh3lvr

Cygbob.
Some more questions.
Is the NH3 inlet point at the lowest part of the shell and is the NH3 exit point and the highest part of the shell and is the oil cooler mounted with a slope(NH3 exit higher than inlet)? Is the return line sloping upwards the whole way back with no traps. Any vapour must have a "free" return back to the system. Is the liquid vessel feeding the oil cooler mounted 2-3 meters above the oil cooler to give the liquid NH3 suffiecent head. A sketch of the system showing the NH3 circuiting around the oil cooler would be most useful. 300 kW of heat sound like a very large compressor, what sort of system is it.

Mike W

Quote:

---

Originally Posted by NH3LVR

Cygbob
A few questions.
Did it operate satisfactorily at first?
Is the Oil Cooler Horizontal or Vertical?
Pardon me if I misunderstood but are there 717 Tubes in the heat exchanger?
Do you have a sketch of the Thermosyphon System you could post?
Is there a lot of oil circulating in the plant?
Nh3lvr

---

hi Nh3lvr,
thank you for the question!
the oil cooler is horizontal, it is 717 tubes in oil cooler , just a coincidence.
we found there was something wrong with the oil filter, but not so badly.
I don't know how it works well and the oil temperature difference between oil cooler outlet and inlet is mostly 40 Deg C, but the values is changing, sometimes the inlet temperature is more than 100 Deg C.

best regards

cygbob

Quote:

---

Originally Posted by Mike W

Cygbob.
Some more questions.
Is the NH3 inlet point at the lowest part of the shell and is the NH3 exit point and the highest part of the shell and is the oil cooler mounted with a slope(NH3 exit higher than inlet)? Is the return line sloping upwards the whole way back with no traps. Any vapour must have a "free" return back to the system. Is the liquid vessel feeding the oil cooler mounted 2-3 meters above the oil cooler to give the liquid NH3 suffiecent head. A sketch of the system showing the NH3 circuiting around the oil cooler would be most useful. 300 kW of heat sound like a very large compressor, what sort of system is it.

Mike W

---

hi Mike W,
thank you!
NH3 inlet point and the exit point has a 3 deg slope, the return pipe has several elbows but the pipe continuously ascend. The liquid vessel is 2.7 meters above the oil cooler.
The ammonia system is designed for a chemical plant, it is definitely a large system.

best regards

cygbob

My first question is; why is the oil cooling load so high? Are you using a common oil cooler for multiple compressors?

What are the tube inside diameters & oil cooler shell diameter?

I have worked on some big systems and I have never seen an oil cooler with 717 tubes. Based on this alone, this tells me you are doing something unique that seems to be unconventional.

Quote:

---

Originally Posted by cygbob

...and if the environment temperature becomes higher ,the oil cooler performance will become much worse.

---

This is usually an indication of problems with the liquid ammonia , or, the baffles in the shell & tube heat exchanger bypassing.

I've seen both. If you are sure the baffle clearances are tight inside the shell, then the problem is probably related to the ammonia piping or oil cooler selection.

What are the diameters of the vent line and the pilot receiver?

Quote:

---

Originally Posted by cygbob

...sometimes the inlet temperature is more than 100 Deg C.

---

On most twin screws the upper limit for discharge temperature is about 100 C. The statement above seems to indicate the problem is loss of oil cooling capability.

Cygbob;
I am not that knowledgeable about Thermosyphon Design. Usually there are very few problems, unless there is a design flaw. I have had problems with oil in Vertical Coolers, but yours is horizontal.
Your sketch shows a drain plug. I certainly hope this is indeed a valve. I trust the Liquid Supply comes off the bottom of the vessel.
I am concerned about the number of tubes in the cooler. The tube size could cause a problem.
Thermosyphon is sensitive to Head Pressure. Are you above design?
Since it appears to be a design and not a operational problem I should probably leave you in the hands of USiceman, as he is much more knowledgeable than I in such matters.

Hi, cygbob :)

Quote:

---

Originally Posted by cygbob

300KW, tube counts:717, circulation ratio:3:1, oil inlet/outlet temperature:80/40 Deg C, ammonia: 38/39 Deg C.

---

Here you have dT=1-2 C

Quote:

---

Originally Posted by cygbob

251KW, oil:87.7/49, ammonia:33.3/35.4,

---

And here you have dT= 14-16 C

Are you sure in this values;)

Speaking about your scheme is the venting line from siphon receiver to condenser really connected in that way:confused:

Best regards, Josip :)

Hi just to be sure.Du you have an reciver under the normal reciver for the oil cooler as i think you are running dry of nh3 sometimes.

Best regards
hgs

It would be best if you could make a sketch of the condenser/receiver system with the thermosiphon receiver shown and indicate the relative heights between the components

Normally, I take the ammonia return line from the oil cooler back into the thermosiphon receiver then have a large vapor vent line back to the condenser inlet. Also, the tube side of the cooler must be single pass with large lines in and out of the cooler. Lines need to be open so that they will not trap, leading to being vapor bound or locked.

Since you manage to start out working OK and then get worse, you might check your piping to be sure you are not becoming vapor locked.

Out of courisity, who is the supplier of the equipment?

Ken

Hi all,
The sketch I uploaded yesterday has some mistakes and I uploaded a new one!

It is the first time for us to design thermosiphon oil cooler, so we want somebody experiential to tell us what problems the system had, and whether the heat transfer area is sufficient?

ps: tube:￠12*1, length:3800mm, slick carbon steel.

Quote:

---

Originally Posted by US Iceman

My first question is; why is the oil cooling load so high? Are you using a common oil cooler for multiple compressors?

What are the tube inside diameters & oil cooler shell diameter?

I have worked on some big systems and I have never seen an oil cooler with 717 tubes. Based on this alone, this tells me you are doing something unique that seems to be unconventional.



This is usually an indication of problems with the liquid ammonia , or, the baffles in the shell & tube heat exchanger bypassing.

I've seen both. If you are sure the baffle clearances are tight inside the shell, then the problem is probably related to the ammonia piping or oil cooler selection.

What are the diameters of the vent line and the pilot receiver?



On most twin screws the upper limit for discharge temperature is about 100 C. The statement above seems to indicate the problem is loss of oil cooling capability.

---

hi US Iceman,
It's a single sompressor and the oil cooler heat load is 300KW:)
tube inside diameters /shell diameter: 10mm/460mm;
Do you think the heat transfer area is sufficient? Or any other problem in our design??

Quote:

---

Originally Posted by NH3LVR

Cygbob;
I am not that knowledgeable about Thermosyphon Design. Usually there are very few problems, unless there is a design flaw. I have had problems with oil in Vertical Coolers, but yours is horizontal.
Your sketch shows a drain plug. I certainly hope this is indeed a valve. I trust the Liquid Supply comes off the bottom of the vessel.
I am concerned about the number of tubes in the cooler. The tube size could cause a problem.
Thermosyphon is sensitive to Head Pressure. Are you above design?
Since it appears to be a design and not a operational problem I should probably leave you in the hands of USiceman, as he is much more knowledgeable than I in such matters.

---

Hi,NH3LVR:)
Thank you very much, you are so kind.

Best regards
cygbob

Quote:

---

Originally Posted by Josip

Hi, cygbob :)



Here you have dT=1-2 C



And here you have dT= 14-16 C

Are you sure in this values;)

Speaking about your scheme is the venting line from siphon receiver to condenser really connected in that way:confused:

Best regards, Josip :)

---

Hi Josip,
Thnak you! Your question is very important, I don't why, but the values are measured a few days ago, I think except some measured errer they should be correct.

Another problem is that the fulid out of the condenser has a 9℃ sub-cooling! There must be something wrong with the system!

Best regards
cygbob

What is the diameter of the vent line pipe? And, what are the dimensions of the pilot receiver?

You have a lot of subcooling! This may be affecting the continuous flow needed to keep the Thermosyphon vessel filled.
I recently ran into this problem. We had multiple condensers and plenty of capacity, but high head pressure and high subcooling. The piping for the equalizing line was questionable.
We valved off the inlet to one of the condensers and saw no major improvement. We started the purger (which was not piped well) and the head came down and everything worked fine.

Hi Cybob:)

I will go for the old favorites:D

1. liquid supply trapped going to the oil cooler

2.Incorrectly sized wet return riser from the oil cooler to the condenser

3. Oil trapped in the refrigerant shell.


If you check this and rule them out it will help you reach a valid conclusion:)

Kind Regards Andy:)

Quote:

---

Originally Posted by US Iceman

What is the diameter of the vent line pipe? And, what are the dimensions of the pilot receiver?

---

the diameter of the vent line pipe is DN100,the pilot receiver dimension is about 1 m3

Based on a very quick review, the pipe sizes you have listed appear to be OK. The pilot receiver volume also appears sufficient.

In your last drawing you stated the liquid drain temperature from the condenser is 26.6 C (79.9 F), but after the liquid flows through the receiver the liquid ammonia temperature suddenly increases to 34.4 C (93.9 F).

What is the condensing pressure?

Can you provide a drawing of the internals of the receiver also please?

Quote:

---

Originally Posted by US Iceman

Based on a very quick review, the pipe sizes you have listed appear to be OK. The pilot receiver volume also appears sufficient.

In your last drawing you stated the liquid drain temperature from the condenser is 26.6 C (79.9 F), but after the liquid flows through the receiver the liquid ammonia temperature suddenly increases to 34.4 C (93.9 F).

What is the condensing pressure?

Can you provide a drawing of the internals of the receiver also please?

---

ok! the condensing pressure is 12.7barg. the delivery gauge is about 35.5 ℃. the receiver is empty!
but ,I don't understand why the dT is so large!

best regarsds
cygbob

lol China must have an astronomical amount of refrigeration with all of the process and manufacturing!!! cant even imagine, my thoughts are where is your designer and consultant in this? good luck stan

Quote:

---

Originally Posted by Mike W

Cygbob.
Some more questions.
Is the NH3 inlet point at the lowest part of the shell and is the NH3 exit point and the highest part of the shell and is the oil cooler mounted with a slope(NH3 exit higher than inlet)? Is the return line sloping upwards the whole way back with no traps. Any vapour must have a "free" return back to the system. Is the liquid vessel feeding the oil cooler mounted 2-3 meters above the oil cooler to give the liquid NH3 suffiecent head. A sketch of the system showing the NH3 circuiting around the oil cooler would be most useful. 300 kW of heat sound like a very large compressor, what sort of system is it.

Mike W

---

Hi Mike,

Is there a good piping design manual available for NH3 based thermosiphon oil cooling systems?

Thanks!

Ernie,

Please only post a question or request in one area please. It helps to keep the answers and replies together and also keeps down the posts we review.

It would be a big help. Thanks in advance.

To go back to basics,is there air trapped in condensor/thermosyphon area causing flow and apparent subcooling at times?
NH3LVR has mentioned this .
Also is condensing pressure constant?
Sometimes if condensing pressure lowers to fast it can cause bubbles in t/syphon circuit.
If liquid drain into pilot reciever from condensor is directly above liquid feed to oil cooler it may sometimes break t/syphon seal.This liquid outlet probably should have venturi eliminator.
US Iceman may also shed light on liquid drain line from condensor to reciever.M aybe it should have liquid trap, unless it dips into pilot reciever level?

Sorry i got excited and didnt look at date.

Quote:

---

Originally Posted by Ranger1

This liquid outlet probably should have venturi eliminator.

---

I'm assuming you mean vortex eliminator because of what you were saying. And yes, I would agree with you. In fact, it's a great point to bring up as a lot of people only consider this for pump suction lines and would never consider this for a simple little vessel.

Piping is a big issue with thermosyphons and gravity drainage from condensers. Just look at how much trouble we have with liquid hanging up in condensers now.:D

If the piping is not close to 100% perfect something almost always will not work correctly.

If thermo oil cooler is not working and was before, there is not enough liquid in receiver to feed oil cooler.
magoo

Thermosyphon design is about 1/2 science and 50% black art (witch craft). I would be concerned about the small diameter of the tubes. Generally t-siphon coolers will do much better with tube diameters of 19 or 25 mm. I believe you said the cooler has a tube length of 3.8m? You may be creating a gas locked condition if you are vaporizing the nearly all of the refrigerant.

DO upi have any idea of the quality of the wet return? If it is running warmer than the inlet, you may be coming close to gas bounding the cooler.

A patch for this without replacing the cooler is to get a canned turbo-regenative pump and pump the liquid through the cooler to insure adequate liquid feed.

I have never seen what I would call a cook-book thermosyphon cooler design manual. It is something that most of us have collected over time.

Ken

Ken

Quote:

---

Originally Posted by TXiceman

Thermosyphon design is about 1/2 science and 50% black art (witch craft).

---

Isn't that the truth? I don't think I have ever done so much research on a single subject in my life as what was spent on this topic.

Quote:

---

Originally Posted by US Iceman

Isn't that the truth? I don't think I have ever done so much research on a single subject in my life as what was spent on this topic.

---

And when you finished researching....you wondered if you knew any more than when you started.

Some of the things I learned over the years were:
-keep line sizes plenty big.
-minimize fittings
-minimize or eliminate valves. If you have to use a valve, a full ported ball valve is preferred or a large ported angle globe valve and be very careful where you place the valves.
-Keep tube diameters on the larger side and keep them as short as is practical
-never use more than a single pass on the tube side
-if practical, slant the tubes up to ward the outlet.
-put and oil drain connection on the inlet head of the cooler.
-double check your pressure drop calculations and insure you have plenty of static head.

All of this and it will PROBABLY work...but no guarantees.

For as many years as T-syphon coolers and evaporators have been used, you would think that there would be more written on the subject that is based on scientific studies.

I think he has a big problem in the 10mm diameter tubes. The are only 0.393 inches in diameter. I have never seen any under 19mm diameter used.

I wonder if he has anyway of accurately getting a pressure drop across the tube side to see if it is anywhere near the predicted dP and how it reacts with load (flow).

Ken

Quote:

---

Originally Posted by US Iceman

Ernie,

Please only post a question or request in one area please. It helps to keep the answers and replies together and also keeps down the posts we review.

It would be a big help. Thanks in advance.

---

Sorry about that.

TXiceman, interesting about tube size as we only use 3/8 " tube on all water and t/syphon shell and tube heat exchangers without any problems.
I think a lot of australian info on design was european technology from stal.

Quote:

---

Originally Posted by Ranger1

...info on design was European technology from stal

---

If there is any way of getting some links or copies of this material I would be very grateful. This is one of those topics where all reference material is worthwhile reviewing for other ideas and past experience.

My comments would match those of TX iceman also. The smallest heat exchanger tubing I have used is 0.75" with a 0.035" wall thickness. No problems encountered on that size or larger.

The biggest problems I have seen are the piping installations and the ability to vent gas or to supply liquid, which almost always goes back to the piping design and pipe sizes selected.

Its out of my league, maybe someone with a bit more engineering expertise can help.We dont even slope oil cooler and it seems to work!

A single pass oil cooler placed in a horizontal plane will work OK. The oil coolers with a slight slope upwards on the outlet seem to work just a bit better because they vent gas so much better.

I just remembered a project I worked on with propane. The oil cooler had 3/8" tubes and it seemed to work OK. The difficult task is to quantify what OK means though. I don't remember the oil coolers having any difficulty meeting the heat rejection requirements, so I'm assuming it was acceptable.:o

I have, somewhere, a Sabroe technical data bulletin on thermosyphon cooling used for the SAB screws. I'll root around for it amongst the cases I've packed for moving house if you would like me to and have a look at what they recommend.

We had problems with our SAB's on start up's and on sunny summer early afternoons (when the sun was full on the HP receiver). Our pipe runs were long - too long really - and although it checked out on paper - I'm pretty sure it was down to lack of mass-flow to shift the heat away quick enough. Once it tripped on oil temp, we fought a losing battle until the sun moved off the condenser platform.:(

A pumped system may help but you still need sufficient suction head (say at least a meter drop off the receiver) to prevent cavitation on some pump types.

Steve

Quote:

---

Originally Posted by Steve Dixey

We had problems with our SAB's on start up's and on sunny summer early afternoons (when the sun was full on the HP receiver). Our pipe runs were long - too long really - and although it checked out on paper - I'm pretty sure it was down to lack of mass-flow to shift the heat away quick enough. Once it tripped on oil temp, we fought a losing battle until the sun moved off the condenser platform.:(

---

That sounds like the solar heat gain generated too much vapor for the receiver equalizing line to vent back to the condenser inlets. I'll bet if you insulate the receiver the problem would greatly disappear.

If you could find that bulletin and get it transformed into a pdf I'm sure a lot of people would find it benefical Steve.

Quote:

---

Originally Posted by US Iceman

That sounds like the solar heat gain generated too much vapor for the receiver equalizing line to vent back to the condenser inlets. I'll bet if you insulate the receiver the problem would greatly disappear.

If you could find that bulletin and get it transformed into a pdf I'm sure a lot of people would find it benefical Steve.

---

We tried all sorts, including spraying the receiver with water, but we never really solved it, and by then we'd been taken over by a bunch of "non-spending, run it till it dies", accountants.

The equalizing line was a DN100 for a 1400kW system so it seemed well sized and was a very direct route.

I'll dig the tech paper out then. Don't know about copyright issues though....

Steve

Quote:

---

Originally Posted by Steve Dixey

We tried all sorts, including spraying the receiver with water, but we never really solved it,...

---

A DN100 is a decent size, however...if the oil temps are still shutting down the screws and all of the other usual suspects have been identified (oil cooler sizing, compressor duty, etc) the probability of excess vapor not venting adequately can usually be tied to the vent line. These pipes have to be large to allow very low pressure losses for the vapor flow they can experience.

Quote:

---

Originally Posted by Steve Dixey

... and by then we'd been taken over by a bunch of "non-spending, run it till it dies", accountants.

---

Arrrgghh! The bane of all honest guys trying to do a good job. These are the same people who complain when something breaks down because they would not provide the money to prevent it.;)