SOLUTIONS TO OVERCOME LIQUID ENTRAINMENT IN SUCTION LINE 1

[allan01]

Q What industry vapour velocities are commonly employed in liquid vapour separator vessels. I specifically refer to an industrial ammonia refrigeration plant. A particular plant has a high stage liquid accumulator/surge vessel/intercooler from which 6 single stage high stage ammonia recip compressors take suction. The low stage and high stage load is mostly constant. The 2 compressors with suction droppers closest to vessel suction nozzle have historically suffered most with liquid carry-over.But I also suspect a build up liquid condensation in the suction droppers which are unusally long,probably due original designers poor knowledge of pump requirements. Seperation velocities that are being employed according to myself are between 0.1m/s to 0.5m/s(-10 to -25 deg Centigrade)I have not yet had an opportunity to look at the process design for the vessel or look at vessel internals yet. I also have some suspicion that the capacity control configuration or anti-recycle time has some contribution as well(specifically to the condensation in suction droppers).
Q Are demister pads in vessels in this service an accepted norm ? I assume with this would come a manway and regular internals inspection.
Q Would an in-line heat exchanger type flanged spool with channel heads placed in the suction header subcooling +35 degC liquid ammonia assist with vapourizing any liquid droplets from the gas stream(naturally optimizing thermal length not to be penalized on superheating). This would be completely unconventional (out of the box thinking)for a system of this size.
Q What thoughts does anyone have on maintaining the wall of the suction droppers above the dew point(saturation temp) of the high stage vessel pressure to eliminate condensation in suction dropper.


I look forward to network with specialists or similar industries.

ALLAN01

 

[Montemayor]

ALLAN01:

What I think you are trying to describe is a conventional Ammonia Flash Intercooler vessel as used in a 2-stage Ammonia mechanical compression cycle. This type of intercooler injects the discharge of the 1st stage Ammonia compressor(s) into the vessel through a dip pipe, effectively sparging the hot discharge gas into a bath of cold, saturated liquid Ammonia produced by flashing a portion of high-pressure NH3 liquid into the vessel. The resulting, cooled 1st stage discharge gases and the evaporated gas are taken up by the 2nd stage NH3 ammonia compressor(s) and compressed to produce the high-pressure, condensed liquid that is eventually used as the source for liquid NH3 expansion into the 1st stage flash intercooler and the main NH3 refrigeration evaporator.

Some NH3 intercoolers employ a closed coil for the hot 1st stage discharge instead of openly sparging it into the cold liquid bath. The coil in installed submerged in the cold NH3 liquid within the separator vessel and exits the separator and into the 2nd stage of compression. The corresponding liquid NH3 boil-off caused by the cooling joins the cooled 2nd stage suction vapors and is also recompressed there. I’ve used both varieties and consider the direct sparging type to be more efficient. Other vapor-liquid separators are employed when one uses a recirculated (or pumped) cold liquid NH3 stream to various cold coils – such as often used in the food processing industry. The basis for the design is the same: the well-known Souders-Brown correlation is used to identify the maximum allowable superficial vapor velocity and from this calculated number, the dimensions of the separator vessel are derived. The vessel sizing is discussed in thread124-125547 and thread124-125878 among others in these fora. Note that you have to calculate the maximum allowable superficial velocity – you don’t apply an assumed one. This velocity is dependent on the fluid’s (Ammonia’s) vapor and liquid density at the separation temperature – a value you can easily and accurately obtain from:

http://webbook.nist.gov/chemistry/fluid/.

You can also employ demister pads internal to the separator vessel. I never apply demister pads or internal parts in vessels for this type of service. The obvious maintenance and inspection requirements make it troublesome and a hazardous operation, especially in NH3 service. My separators have operated for years and years without internals, problems, or entrainment of liquid particles. By not relying on internals and keeping the design simple, I eliminate the need for inspections and hazardous vessel entry in the future.

You could also employ a heat exchanger on the hot gas. I’ve discussed this above, using the coil in the NH3 liquid bath. This is a more expensive version of the same intercooler. As I said, I’ve used both; the simpler sparging type has always given the desired results without problems and the simpler I keep a process, the less the maintenance and the problems of instrumentation, process upsets and capital costs.

If you employ the Souders-Brown correlation and generous vapor disengagement space above the feed flash nozzle, you will have no problems (like me) in having liquid entrainment exiting the separator in the overheads vapor stream – just keep the vapor superficial velocity low.

This same design criteria is employed in designing the suction drum to the 1st stage of compression in any refrigeration compressor sucking the vapors from a refrigeration evaporator. The principle is the same.

I hope this helps you out.