Water as a buffer fluid for ammonia mechanical seal

[Mechanic]

All,
Is anybody aware of using water as a buffer fluid in mechanical shaft seal for ammonia service?
It is a new design of the turbine working on gaseous ammonia in close energy recovery cycle and not in refrigeration cycle. Pressure and temperature at seal inlet are 125 psi and 70 deg F, speed 1800 rpm and shaft DIA 6". Big machine.
With stand alone oil bearings (magnetic bearings in the future) in mind it looks like expensive dry gas seal plus air/ammonia or nitrogen/ammonia separation system are the only choice. In the same time with use of water as a buffer I can probably use much less expensive, more simple face seal with all needed arrangements of clean ammonia vent, ammonia/water mix vent and air/water mix vent.
I am new in the ammonia field and not sure about ammonia/water separation system. How complicated and expensive it is, how clean will be ammonia after separation and can it be return to the cycle. Please help. Thank you in advance.

[RANGER1]

Hi Mechanic,
Water may be OK but you would nor want it to cold.
Will it be in a jacket around the seal?

In older machines rotary vane compressors they used water or transformer oil.
The water system was used with the plants cooling tower, usually with thermostatic water valve to maintain temp.
Transformer jacket cooling was used in conjunction with a shell & tube heat exchanger, water cooled.


On your system you could possibly mount a small vessel at a calculated height & size, volume & use thermsyphon method. Ambient would naturally cool circulating oil in seal jacket.

[Mechanic]

Hi RANGER1,

The lowest temperature in the cycle is 13.5 deg C (condenser temperature at 6.2 bar pressure). So there is no danger to freeze the water.
With single face seal water will be injected into the seal (not into a seal jacket) behind face sealing rings at slightly lower pressure than ammonia pressure in the seal cavity. It will first go around sealing faces and cool them, next it will go out from the seal with small amount of ammonia that leaked through the sealing faces. With more complicated double seal principle will be the same small amount of ammonia mixes with water and vent from the seal.
Here where we need separate ammonia from the water and return it into the process. You mentioned small vessel with thermosyphon cooling. Is it intended to be an ammonia/water separation device? can you explain how it works in more details. How clean will be ammonia after separation? If it is not absolutely free of water how water may be removed from the process. Any references to show managers will help a lot.

[RANGER1]

Hi RANGER1,

The lowest temperature in the cycle is 13.5 deg C (condenser temperature at 6.2 bar pressure). So there is no danger to freeze the water.
With single face seal water will be injected into the seal (not into a seal jacket) behind face sealing rings at slightly lower pressure than ammonia pressure in the seal cavity. It will first go around sealing faces and cool them, next it will go out from the seal with small amount of ammonia that leaked through the sealing faces. With more complicated double seal principle will be the same small amount of ammonia mixes with water and vent from the seal.
Here where we need separate ammonia from the water and return it into the process. You mentioned small vessel with thermosyphon cooling. Is it intended to be an ammonia/water separation device? can you explain how it works in more details. How clean will be ammonia after separation? If it is not absolutely free of water how water may be removed from the process. Any references to show managers will help a lot.

Mechanic,
The thermosyphon oil cooling system I suggest is a closed circuit system which circulate naturally.
This probably is not suited for your application.

If you use water to cool seal, are you concerned about fouling seal area, corrosion etc.

I can only think of a cartridge or similar seal cooled with water or oil but pressurized above plant pressure with nitrogen so eg nitrogen leaks into your system then is removed by a automatic refrigerated purger.

This stops anything except nitrogen leaking to atmoshere.

I think your best bet is to get a good seal company to work out a solution.

I can try to dig up something on thermsyphon method!

[sterl]

http://www.johncrane.com/PDFs/B_air.pdf

Separating water from ammonia for refrigeration is pretty easy, stills are made by HA Phillips and RVS and a host of others. They do rely on lower suction pressures than you are dealing with but they typically can reduce water content to about 0.1% operating at something just past 120-deg. aqua and suction (of plant) near atmospheric. There is an IIAR bulletin that has a pretty self explanatory chart of partial pressures and concentrations, etc.

Is the space containing the compressor terribly sensitive to the ammonia vapor? I would think you would have considerably better luck using an oil in a closed loop arrangement as a seal barrier fluid than water; especially if there are any contaminants in the ammonia gas stream.....Check out Cornell refrigerant pumps and their seal fluid circulation arrangement (double mech, the pump is intended to operate below atmospheric at suction or in standby....). This mechanism a little sluggish in response to quick changes in pump discharge pressure.

If you need to utilize water I would be looking at something like the API 54 arrangement.