Strange drop in compressor flow

[aakbar]

Hi everyone,

We are facing a strange issue. After we start up the system we notice quite a lot of drop in compressor discharge flow. Simultaneously the suction pressure drops and temps go down with it. This continues for a while...15-30 minutes. And then suddenly as it appeared, the flow goes back up, pressures return to normal and the temps go back up as well. Several times we've had to turn the system off to prevent any accidents.

My only hypothesis is that somehow we end up in a situation where there's not enough vapor in the liquid pressure receiver which causes low suction pressure and low flow. But thats just a hypothesis and I am not an engineer. Just the owner.

I am attaching pictures of the computer screen for reference. The solid blue line is the evaporator inlet Temp and black line is the outlet. Ignore some of the pressure reading being non-sensical. We're undergoing some PT replacements.

If anyone can help in this matter, I would greatly appreciate it.

Regards

PS: forgot to add: the plant is located in Pakistan and we manufacture block ice. We are using ammonia gravity feed.


Screenshot 2023-11-28 at 6.29.47 PM.jpgScreenshot 2023-11-28 at 6.33.24 PM.jpg

[seanf]

Are the ammonia liquid levels ok in the low pressure receiver? and high pressure receiver if it has one.
Is there any noticable liquid level changes during that drop in flow?

Is the system still the same one from back in 2016?

[aakbar]

Are the ammonia liquid levels ok in the low pressure receiver? and high pressure receiver if it has one.
Is there any noticable liquid level changes during that drop in flow?

Is the system still the same one from back in 2016?

Hi seanf, we switched from direct expansion to gravity feed around 2018.

There is a slight bump in LP liquid level when the compressor flow drops. But it doesn't seem out of ordinary. We don't have a level sensor in the High pressure receiver. And the odd thing is it just seems to happen right after the start up. Afterwards the system works just fine.

I've attached screenshots for the LP level and the corresponding chart for compressor flow. Please take a look.

Screenshot 2023-12-01 at 10.38.52 AM 1.jpgScreenshot 2023-12-01 at 10.40.54 AM.jpg

[seanf]

Is everything started at the same time (compressor, condensor fan, condenser water pump, brine pump, ice production, ect...) or is there a sequence thats followed?

Is there a LP receiver level column or sight glasses you can check the level sensor against?

For the HP receiver, do you still have the thermal camera?

Is the compressor on a fixed speed motor, and does it have any capacity control?

Does the condensor fan have any speed or cycle control, or is it ran full speed all the time?

[seanf]

Does your chart let you look at data from say, a year ago?
Sometimes its handy to look at trends from times when conditions were similar (ambient temp ect..), or to see if theres been a slow change in system operating conditions over a period of time.

[aakbar]

We follow this sequence at start up:
Condenser pumps/fans come on first.
Then brine agitator comes on
Then the compressor is started unloaded. After a minute or so the operator slowly opens the suction valve.

We do have a gauge glass on LP. I can ask the operator to make a video of HP and LP gauge glasses when he notices a drop in suction pressure.

I am back in the US for a few months currently and the thermal cam is sitting here too.

All motors are fixed speed and compressor does not have capacity control. It is a 2 cylinder 9"x9"x300RPM compressor.

The PLC/Scada is a recent plant upgrade. So I only have a few days worth of data. I haven't figured out a reasonable historian yet. Everything costs too much to justify in a small plant like this so my plan is to eventually figure out a way to send data from plc to sql file and then to grafana. Someday.

[seanf]

It makes me think that initially either theres not enough liquid being supplied to the LP receiver, or that theres something restricting the liquid ammonia flow in the evaporator.

I would be looking at the LP and HP gauge glasses to see whats going on with the liquid levels. Maybe its getting hung up in the condenser. Seen a few liquid level sensor give funny readings.

But also thinking about, does the evaporator need drained of oil. Maybe theres a little bit in there that shifts/spreads out when the pressure drops.

You could always make a few paper records at different times in the year, handy to have something to compare too.

[cricri]

Hello aakbar,
your starting sequence is unusual for me.
you should start compressor first and wait the HP to rise (pressure switch, HP sensor) before starting pump and fans on condenser.
If ambiant temperature is high around the liquid receiver, ammonia start to migrate to condenser.
Pressure differencial LP/HP is low when compressor is starting to pump.

[aakbar]

Hello aakbar,
your starting sequence is unusual for me.
you should start compressor first and wait the HP to rise (pressure switch, HP sensor) before starting pump and fans on condenser.
If ambiant temperature is high around the liquid receiver, ammonia start to migrate to condenser.
Pressure differencial LP/HP is low when compressor is starting to pump.

We switched the start up sequence and turned on compressor 5 mins before the condensers and witnessed the same result. You can view the result in the attached screenshot. Screenshot 2023-12-03 at 6.56.43 PM.jpg

I also received a video of the liquid level in the liquid receiver at the time of this drop in flow. Here it is:

https://youtube.com/shorts/8NcYc_c18_s?feature=share


Here is the liquid level when the flow has returned to normal

https://youtube.com/shorts/XjZM7LtGrvo?feature=share

[aakbar]

Other system numbers right now

Screenshot 2023-12-03 at 7.06.35 PM.jpg

Ignore Tank 1/2 evaporator pressures and Delta P values. Those PT's are undergoing replacement.

[RANGER1]

aakbar,
Do you have any level indication on high pressure liquid receiver?

Should be about 20% level when everything running normally.

What type of level control do you have on gravity vessel, float operating solenoid valve?
Is it operational as soon as compressor starts?

[seanf]

Other system numbers right now

Screenshot 2023-12-03 at 7.06.35 PM.jpg

Ignore Tank 1/2 evaporator pressures and Delta P values. Those PT's are undergoing replacement.

Is the anhydrator there because of moisture in the system, or a just in case sort of thing?

[aakbar]

aakbar,
Do you have any level indication on high pressure liquid receiver?

Should be about 20% level when everything running normally.

What type of level control do you have on gravity vessel, float operating solenoid valve?
Is it operational as soon as compressor starts?

We have gauge glass on HPR. Here is what it showed at the time of these fluctuations.

WhatsApp Image 2023-12-03 at 18.43.37.jpeg

We have a danfoss electronic expansion valve. It opens/closes based on the liquid level in the low pressure receiver.

[aakbar]

The anhydrator and OS1 bit aren't operational yet. When i was working on the HMI, I decided to add everything we will eventually have on the screen so that we're not rearranging the icons later on. Currently we drain water from the LPR manually.

[NH3LVR]

Have not been following this closely, but are you draining water from the low pressure receiver? Perhaps we have the terms confused.

[aakbar]

Yeah..we drain it from the low pressure receiver. Its the vessel that receives refrigerant from the High pressure receiver. There's an expansion valve on the line between these two vessels. It also supplies and receives refrigerant from the evaporators and provides gas to compressors for suction.

[NH3LVR]

That is an LPR. So you do have a separate vessel underneath the LPR (usually called an oil pot) that you drain from? How much water are you getting? Is the system running in a vacuum?

[Tycho]

Are you draining any oil from the evaporators?

Seeing as the evaporators are the lowest point on the LP side, some oil is bound to move from the HPR to the LPR, and it will collect in the evaporators.

I'm guessing the return lines from the evaporators go to the top of the LPR kinda like this:
Screenshot_244.jpg

an oil plug in one or both of the evaporators preventing free flow of liquid could explain the sudden drop in suction pressure and lack of gas flow in the compressor.

if the return lines on the evaporators are piped in on the side of the LPR, they should at least be above the liquid level, another case could be if the liquid level in the LPR increases to go above the evaporator return line and prevents free flow of evaporated gas back up to the LPR.

[aakbar]

We don't have a separate vessel. I've attached picture of oil drain at the bottom of the vessel. I'll ask the operator to drain it of liquid and send the video. It seems to catch quite a lot of water. Thats why we decided to add an anhydrator.

Screenshot 2023-12-04 at 11.23.32 AM.jpg

System is not running a vacuum. At shut down, operator turns off the valve feeding liquid to the low pressure receiver. We turn off the compressor when it reached 1bar suction.

[aakbar]

We drain it every winter. I'll see if I can find any videos from last year..but the entire operation is quite difficult so we only do it once a year. It requires vacuuming the evaporators, connected lines, disconnecting flanges, lifting the coils, tipping them etc..A whole song and dance. A lot of oil does accumulate there though. However, if there is oil in the evaporator, why does the system go back to normal? Why does this hiccup only occur 15-20 mins after start up. Its really strange for me.

[aakbar]

Are you draining any oil from the evaporators?

Seeing as the evaporators are the lowest point on the LP side, some oil is bound to move from the HPR to the LPR, and it will collect in the evaporators.

I'm guessing the return lines from the evaporators go to the top of the LPR kinda like this:
Screenshot_244.jpg

an oil plug in one or both of the evaporators preventing free flow of liquid could explain the sudden drop in suction pressure and lack of gas flow in the compressor.

if the return lines on the evaporators are piped in on the side of the LPR, they should at least be above the liquid level, another case could be if the liquid level in the LPR increases to go above the evaporator return line and prevents free flow of evaporated gas back up to the LPR.

They are connected the way you've shown in your diagram. I've drawn it with bottom connections for connivence.Screenshot 2023-12-04 at 11.32.13 AM.jpg

[aakbar]

Video of water/oil mix from the LPR:

https://youtube.com/shorts/-hjzZlq6R2A


It is not recent. But it is for the system in question

Attachment 16297

[Tycho]

They are connected the way you've shown in your diagram. I've drawn it with bottom connections for connivence.Screenshot 2023-12-04 at 11.32.13 AM.jpg

Since it only happens at start up, I would say it's a a plug or restriction in the liquid supply to the evaporator coils.

as in oil has collected in the bottom of the liquid supply, when you start up, all or some of the ammonia in the coil will boil off and after a while liquid pushing down on the "plug/obstruction" will push through and everything will function as normal.

That's why I were asking if there was a drain point on the evaporator, not only on the LPR.

Don't know what kind of evaporators you have, but I guess the liquid is entering on the bottom and then passing upward through the coil.

if that made sense

Screenshot_245.jpg

[RANGER1]

We don't have a separate vessel. I've attached picture of oil drain at the bottom of the vessel. I'll ask the operator to drain it of liquid and send the video. It seems to catch quite a lot of water. Thats why we decided to add an anhydrator.

Screenshot 2023-12-04 at 11.23.32 AM.jpg

System is not running a vacuum. At shut down, operator turns off the valve feeding liquid to the low pressure receiver. We turn off the compressor when it reached 1bar suction.

I would suggest do not pump down to 1 bar, just turn off, close compressor suction if you must.

If you have compressor crankcase heater it can be left open, this is really old school operation, but you may have another reason?

[NH3LVR]

What kind of compressors do you have? How much oil are they putting into the system?

[aakbar]

Since it only happens at start up, I would say it's a a plug or restriction in the liquid supply to the evaporator coils.

as in oil has collected in the bottom of the liquid supply, when you start up, all or some of the ammonia in the coil will boil off and after a while liquid pushing down on the "plug/obstruction" will push through and everything will function as normal.

That's why I were asking if there was a drain point on the evaporator, not only on the LPR.

Don't know what kind of evaporators you have, but I guess the liquid is entering on the bottom and then passing upward through the coil.

if that made sense

Screenshot_245.jpg

Your suggestion makes sense. We will drain the evaporator and I will report back the results.
I got hold of oil drain videos from last year.

https://www.youtube.com/shorts/GwgD9zuaVcM
We drained 3 buckets from each bank of the coil.

Below is a drawing of the evaporator in question. It is a real pain to drain it of oil. So its an annual routine. Usually in the dead of winter.
WhatsApp Image 2023-12-04 at 19.30.24.jpeg

Here's LPR oil drain: https://www.youtube.com/shorts/-hjzZlq6R2A

[aakbar]

I would suggest do not pump down to 1 bar, just turn off, close compressor suction if you must.

If you have compressor crankcase heater it can be left open, this is really old school operation, but you may have another reason?

I can talk to the operators about it. I think vacuuming might be helpful for unloaded start but its a question I've never asked.

[aakbar]

What kind of compressors do you have? How much oil are they putting into the system?

The current one is very old..probably from 50's or 60's. It matches closely the compressors shown in Hand of Mechanical Refrigeration book from 1922. It was there when I took over in 2016. Its going to be on standby duty soon. We're putting in a Kirloskar TX4 next to it.

[RANGER1]

I can talk to the operators about it. I think vacuuming might be helpful for unloaded start but its a question I've never asked.

I think when you pump down to 1bar you are lowering level in gravity feed vessel, it drops level so it does not work on restart.

Turn off at normal running suction pressure.
Close compressor suction valve.
On restarts you MAY have to keep suction stop valve restricted, open gradually bringing suction pressure down.

[RANGER1]

The current one is very old..probably from 50's or 60's. It matches closely the compressors shown in Hand of Mechanical Refrigeration book from 1922. It was there when I took over in 2016. Its going to be on standby duty soon. We're putting in a Kirloskar TX4 next to it.

Make sure it has a good oil separator to avoid oil carry over.
Instal an oil drain pot on bottom of gravity separator

[NH3LVR]

First off my concern was the water. It used to be common here that companies would use Agricultural Ammonia. It contained water of unknown quantities.
I think RANGER1 is spot on with his recommendation to install an Oil Pot. If you open it and let it fill, then valve it off and open a suction line to it you can easily remove the oil before it get to the system.

[RANGER1]

By pumping down to 1 bar you are nearly emptying all ammonia out.
On start up you have to fill it up.
How long to pump down to 1 bar is minimum time to fill back up again.

[Tycho]

Didn't catch the pumping down to 1 bar when shutting down.

But I'm still leaning more towards a plug of oil in the bottom of the evaporator coil, preventing self circulation to start until there is Higher level on the feed side than on the boil of side.

Consider... if the plant was pumped down and liquid was moved to the HPR.

When they started the plant up again, the "pilot valve" would be full open, dumping HP liquid into the LPR at a rapid rate, and the the amount of liquid that is fed in, should expand into a much larger volume than the compressors can take, until the system is in balance.

So instead of the compressor pumping down, they should have to throttle the suction valves on the compressor to keep the gas flow at a manageable flow.


What I see is that they only pump down every now and then, and they only drain oil once a year or so.

Normally, draining oil on an ammonia plant is a weekly part of maintenance.


@aakbar, says they are draining water and ammonia, but this is a positive pressure system, so unless they have received contaminated ammonia in the first place, what they are draining is Oil.

The video Aakbar posted that showed oil drain, that is what oil mixed with ammonia looks like, gray and comes out like toothpaste

Then the video that shows the amount of oil that is drained from the coils... that is a huge amount.

and, as Ranger1 said, I would add an oil collector, not on the LPR, but on the liquid side on the coils.


I mean in HPR recievers that we design today, the pipes stick 10-15CM up into the receiver, and the Oil collection points are flush, to collect the oil.

Not so on older designs, on those, most connections are flush, so oil has to be collected "from the lowest point on the system" and in this case that is the evaporators.


But also, don't pump down... just hit stop on the compressor and let the liquid in the plant settle in whatever receiver

[Peter_1]

Not very helpful but this is a very interesting post for me to read