Precaution in Liquid ammonia Pumps operations

[smpsmp45]

Based on our experience in the recently commissioned plant, I would like to have experts opinion.

It is a plant with Liquid Ammonia Pumps. We have seen that if the plant is shut down for 2 / 3 days for some work / repairs etc (as commissioning is underway), the pumps show erratic behaviour once we restart the plant. It must be due to the fact that the casing & other parts now being exposed to ambient conditions, create hot zone & cavitation takes place.
What could be the simple solution to this cause?

[nh3wizard]

What kind of pumps do you have? Does the pump have a heater?

Could be an oil problem.

[NH3LVR]

We have seen that if theplant is shut down for 2 / 3 days for some work / repairs etc (as commissioning is underway), the pumps show erratic behaviour once we restart the plant. It must be due to the fact that the casing & other parts now being exposed to ambient conditions, create hot zone & cavitation takes place.
What could be the simple solution to this cause?

The simplest solution to cavitation problems with Liquid Pumps is not to shut the system down.
That not always being practical I can give you a few hints.
One cause of cavitation on startup is bubbles in the pump vessel. This is the result of the liquid boiling as the pressure is reduced. I generally try to bring the LPR down to temp before starting the pump.
Oil can cause a problem if there is a buildup in the vessel. When running the oil circulates through the system. When you shut down the pumps oil collects in the lowest place-the pumps. This can take hours to clear out, as it is extremely difficult to remove cold oil from the pumps. If your plant is new this is most likely not your problem.
Another trick I often use is to pull the LPR down to temp, then unload the compressor and allow the Suction to rise a few pounds. This makes the liquid subcooled. After the pumps are running I will allow the compressors to load. Works almost every time.
Another thing that can be done is to reduce the flow from the pumps when they are started. I generally do this by restricting the outlet valve on the pump. This reduces the head required to prevent cavitation.
What kind of pumps do you have?

[US Iceman]

Another trick I often use is to pull the LPR down to temp, then unload the compressor and allow the Suction to rise a few pounds. This makes the liquid subcooled. After the pumps are running I will allow the compressors to load. Works almost every time.

Shhhh. You are giving away secrets...

The solution to starting refrigerant pumps is to have the liquid cold before the pumps are started. It normally takes several hours for the pumps to cool down before starting them.

The next most important item to check is the pump itself. If you are using Qmin & Qmax orifices you should be OK. If not, then you need to make sure the pump discharge head is where it is supposed to be. If the resistance in the liquid line is lower than estimated, or if the pump was selected with too much head capability, the pump can move a lot more liquid which causes the NPSHR to increase drastically.

You need to make sure the pump is commissioned properly.

The last point is; don't allow the suction pressure to decrease rapidly. As the suction pressure drops, boiling in the liquid will cause some of the gas bubbles to get sucked into the pump, which induces cavitation.

Another concern is this; if the system will be shut down for any period of time you need to ensure there are hydrostatic liquid relief valves in the pumped liquid line. When the system is shut off the liquid pressure can increase very quickly in this line because of the check valves and solenoid valves in the main liquid line.

[NH3LVR]

Another concern is this; if the system will be shut down for any period of time you need to ensure there are hydrostatic liquid relief valves in the pumped liquid line. When the system is shut off the liquid pressure can increase very quickly in this line because of the check valves and solenoid valves in the main liquid line.

Very good point!

[Peter_1]

Another concern is this; if the system will be shut down for any period of time you need to ensure there are hydrostatic liquid relief valves in the pumped liquid line. When the system is shut off the liquid pressure can increase very quickly in this line because of the check valves and solenoid valves in the main liquid line
and as a result an exploded pump housing.

[Josip]

Hi, smpsmp45

Sorry about your problem, but for sure someone miss to made start up of the pump/s and plant by the book...

Another concern is this; if the system will be shut down for any period of time you need to ensure there are hydrostatic liquid relief valves in the pumped liquid line. When the system is shut off the liquid pressure can increase very quickly in this line because of the check valves and solenoid valves in the main liquid line.

And you can see at attached photos what happen if someone forget to install or open or close some valve...

It was a problem with design, but much more with ignorant operators closing all valves without any clue what can happen when you trapped low temperature liquid....unfortunately they found it out after about 1 (one) hour....very fast learning

.....that happen on the last day of partial start up of the plant (hardening tunnel for ice cream)....... successful start up of the plant and later on (in the evening) all of us went to celebrate that - nice (bitter) dinner in restaurant.......due to excessive liquid pressure one of expansion joints burst and quite big amount of ammonia escape into air....fortunately no one were there....

No need to describe problems with client, authorities...

Best regards, Josip

[smpsmp45]

I liked the secret being leaked out!!! Many thanks for the valuable comments

[US Iceman]

Please let us know how this works out for you.

[smpsmp45]

Surely I have already informed the Site persons to check out on this & by next week I shall have more details with practical aspects in it

[Ponca Dave]

Shhhh. You are giving away secrets...

The solution to starting refrigerant pumps is to have the liquid cold before the pumps are started. It normally takes several hours for the pumps to cool down before starting them.

The next most important item to check is the pump itself. If you are using Qmin & Qmax orifices you should be OK. If not, then you need to make sure the pump discharge head is where it is supposed to be. If the resistance in the liquid line is lower than estimated, or if the pump was selected with too much head capability, the pump can move a lot more liquid which causes the NPSHR to increase drastically.

You need to make sure the pump is commissioned properly.

The last point is; don't allow the suction pressure to decrease rapidly. As the suction pressure drops, boiling in the liquid will cause some of the gas bubbles to get sucked into the pump, which induces cavitation.

Another concern is this; if the system will be shut down for any period of time you need to ensure there are hydrostatic liquid relief valves in the pumped liquid line. When the system is shut off the liquid pressure can increase very quickly in this line because of the check valves and solenoid valves in the main liquid line.

Excellent points Ice, I just want to add what I do to prevent Hyraulic lock up.I drill a 1/16" hole in the check valves. eliminates the bypass relief which is usually threaded. if more than one pump is installed the idle one gets a little backflow but it's insignificant.

[Grizzly]

Excellent points Ice, I just want to add what I do to prevent Hydraulic lock up.I drill a 1/16" hole in the check valves. eliminates the bypass relief which is usually threaded. if more than one pump is installed the idle one gets a little back flow but it's insignificant.

And when the 1/16" hole becomes wire drawn and enlarges what happens?
I have yet to come across a liquid pump system that does not have adjustable relief lines to dump surplus pressure. Usually back into the surge drum.
Mind as Josip proves they are not fool proof.
Grizzly

[benji]

hey aah nh3lvr i agree with you, you would also get liquid hammer at your evaps right?

[Oregon Jim]

I just love seeing new and different things. I have never seen expansion joints like that anywhere except in an oil refinery sulfur plant. I had no idea that they were used in ammonia systems... If you could post some link to information about them I would really appreciate it.

Also I was very amused by the middle picture in the above post, showing a man standing on a stack of pallets. The regulatory agencies here would just have a field day writing citations for that safety violation!