Where is all the ammonia coming from?

Hey guys, I have a head scratcher for you.

Overview: I have an arena with a single Mycom 8WB compressor pulling its suction from a flooded shell and tube chiller. The chiller liquid level controller is a Hansen Vari-level with a 20" probe fitted to a 4" column that controls the liquid line solenoid. There is also a mechanical high level cutout.

Problem: When the system is running the liquid level in the chiller is maintaining at the bottom of the top sight glass with the Hansen controller reading 70% and once the system satisfies and the the plant shuts down the liquid level will increase and trip the mechanical high level with a reading of 110% on the controller. The increase from 70% to 110% takes 10 seconds. Where is all this liquid coming from?

Theory: Suspended liquid in the surge drum is released at shutdown. The higher liquid level you maintain in the chiller the more suspended liquid there will be in the surge drum. Even after lowering the liquid level the same problem occurred. The liquid line solenoid is not leaking and has been ruled out as a problem.

Any thoughts?

Thanks in advance!

P&S Do you have any pictures, sometimes they help understand.
General & & control column.
Is there the tell tale ice on high level switch when it turns off.

Also is it a new problem, has it worked in the past?
Guessing the shell & tube has a header type surge drum on top with 2 or more take off's.
Maybe level is way to high in s&t causing problem you suggest.
Someone can correct me, but think level in s&t should only be just covering tube bundle giving enough gas space in it to breathe.
If it can't gas & liquid May surge up into header vessel on top of s&t.
Try running lower level like 50% &see what happens.
Then trial & error on level unless you take end off to inspect tube bundle configuration.

Attachment 11241

I will try to get more photos.
- It is a new problem. It has always worked well in the past.
- Chiller has 2 take offs teed together and a short suction pipe run. (10-15')
- I have lowered the level from 70% down to 40%. The liquid will still increase by as much as 20% but it will not reach the mechanical high level. The liquid is now in the middle of the bottom sight glass. This is how the system is currently running. I have only suppressed the problem which hopefully gives me enough time to resolve it.

Hi passandscore.

I have had many level control issues over the years due to oil contamination of the probe.
Have you thought about draining the column and re-calibrating the probe?
At least you would know your level readings are accurate.
Which would also drain any oil that may be confusing the issue?

Grizzly

If oil is not on probe could also be in leg in bottom of control column, this can cause slow reaction in column.
If it can't be drained at bottom of leg blast some high pressure gas or liquid into column after shutting top control column stop valve.
This may push it back into chiller.
We see an oil pot, but can't see how it is piped at the moment.

Grizzly,

I can relate as oil has contributed to the majority of my probe problems over the years.

The column has been drained and the probe re-calibrated. No change.

It doesn`t seem like a probe problem because the physical liquid level increases in the chiller, tripping the mechanical high level. The chiller level maintains properly during operation and only increases right after it cycles off. I won`t rule out the probe or oil. I would just be surprised.

Hi
my pennies worth . Still could be an oil saturated HX and surge vessel. The oil will be in semi solution and accumulate in surge vessel then drain back in off cycle giving the high level fault.
Suggest pump out HX, and drain oil from vessel after it has warmed up.
I could be totally wrong though, try it anyway.

Hi again,
had a thought about the feed solenoid valve, with cold ambient and off cycle in HX the pressures can become close/ similar, if a pilot/ diaphragm type solenoid valve they will pass liquid. Believe me, been there had that problem. Go to a direct action solenoid valve.

RANGER1,

Hot-gas was injected just like you stated and we noticed a temporary improvement.

Attachment 11242

The orange ribbon indicates that the valve is closed.

Magoo,

I will pump out the chiller and drain the oil pot as you requested.

The system has a direct action solenoid.

Quote:

---

Originally Posted by passandscore

RANGER1,

Hot-gas was injected just like you stated and we noticed a temporary improvement.

Attachment 11242

The orange ribbon indicates that the valve is closed.

---

It appears from picture that bottom of control column may be a oil trap which is not free draining
back into vessel.
Once oil gets in, can't drain out.
Consider welding in a drain valve on the bottom of it, at lowest point.
I cleared one recently with the method suggested, but used high pressure liquid from 56kg ammonia cylinder.

Quote:

---

Originally Posted by RANGER1

It appears from picture that bottom of control column may be a oil trap which is not free draining
back into vessel.
Once oil gets in, can't drain out.
Consider welding in a drain valve on the bottom of it, at lowest point.
I cleared one recently with the method suggested, but used high pressure liquid from 56kg ammonia cylinder.

---

Passandscore,
Is there any update on this one yet?
Hope it's sorted out

RANGER1,

This job has been an ongoing problem. I am no longer the lead technician. I have been keeping informed and so far all that has been done is a hot gas flush of the column, and the lowering of level set points. There is a lot of theory going around. The problem will disappear for a month or so and then reappear. We are trying to determine the commonality of those high level days. ie: is it a cold day.

There is one thing that I have been trying to determine. The vent line from the oil pot and the vent line from the probe column bull head together at a tee and the branch line from that tee goes into the top of the chiller. Perhaps the oil vent is overcoming the column vent and causing problems. Not sure if its possible. Like I said, lots of theory. The intake air for the room is also located directly above the oil pot and anytime the room is occupied the cold air blows right on top of it. Perhaps this is the problem. But then the question becomes: What affect would warm or cold air have on the ammonia that is in the pot. Warm would cause it to boil off but would this cause the oil pot vent pressure to overcome the column pressure?

Once the problem is truly solved I will give an update but until then I expect intermittent high levels to continue.

These vents can be the reason. Oil pot work as thermosyphon. Inside oil pot vent pipe will be mixture of liquid and gas and level will higher than level in the chiller. Liquid can go to the probe column. Can you isolate oil pot to check how it affect the level in the chiller?

2 or 3 thing happening when you are starting system in level controls .
when you starting system the liquid inside of surge drams start to boiling and there is a lot of bobbles and level of liquid you see in level pipe is not right until you system comes to normal
(especial in low temperature )and ice built around the pipe .

also position of the connection pipe for level and size of that is important (should not be oil in level pipe)
also time of opening time of solenoid valve to charging NH3 is very important ((throttle vales )) and normally it is about %25 to %35 of off time.

Quote:

---

Originally Posted by passandscore

Hey guys, I have a head scratcher for you.

Overview: I have an arena with a single Mycom 8WB compressor pulling its suction from a flooded shell and tube chiller. The chiller liquid level controller is a Hansen Vari-level with a 20" probe fitted to a 4" column that controls the liquid line solenoid. There is also a mechanical high level cutout.

Problem: When the system is running the liquid level in the chiller is maintaining at the bottom of the top sight glass with the Hansen controller reading 70% and once the system satisfies and the the plant shuts down the liquid level will increase and trip the mechanical high level with a reading of 110% on the controller. The increase from 70% to 110% takes 10 seconds. Where is all this liquid coming from?

Theory: Suspended liquid in the surge drum is released at shutdown. The higher liquid level you maintain in the chiller the more suspended liquid there will be in the surge drum. Even after lowering the liquid level the same problem occurred. The liquid line solenoid is not leaking and has been ruled out as a problem.

Any thoughts?

Thanks in advance!

---

Passandscore,

I have no time to read the other posts but having in count that 10 seconds is very quick to go from 70% to 110% I don't see another good explanation than the one you tell on theory. Other problems may also be present on the system but I think the main is the resident liquid that is in the surge drum and then when the compressor stop the pressure equalization will by gravity take the liquid back to the chiller.

Is there any problem when you start the compressor with liquid level at 110%? Doesn't the surge drum still provide the safety of the compressor at the startup.

Make the following experiences if possible:

» Work with the compressor for a while before it stops (during 5-10 min) at 50%. This will carry less liquid to the surge drum and probably you see the level will not increase...just to comprove you theory.

» If there is problems with the surge drum at start up why don't cut out the liquid feed solenoid valve about 5 min. before the compressor and the water pumps stop...if eventual the compressor will stop by lower pressure reduce the timing to about 2 min...adjust the necessary timing.