Printable View
Hi all,
Is there a rule of thumb for tolerable LP separator pressure fluctuation during operation and pull down rates to prevent pump cavitation?
I seem to recall 10 psi/minute or maybe it was 10kPa/minute. What are the pros and cons for vertical vs horizontal other than to suit available space? Vertical for recips because of stepped loading?
I'm talking ammonia at -20 to -30 evaporating.
Never thought about kpa/minute but thinking every plant different & hard to say for all.Quote:
Originally Posted by Superfridge
If you have pull down to quick pumps cavitation for sure.
You can also stage things on, not turn a big plant on all at once.
Compressors need to load allowing for enough time for reaction.
There is a number of control strategies like PID loops, so further away from setpoint load/unload pulses bigger for screw slides, closer to set point load/unload pulses less.
Vessel has to be sized to cope with all scenarios of course.
Pump cavitation auto start/stop handy for when pumps cavitate pump stops & restarts after a delay to stabilise miltiple times.
Pump inducer & max flow orifices helpful as well as max flow control valves to keep pump in pump curve.
I thought vertical vessel has more storage capacity/volume for fluctuating levels.
Horizontal keeps things more compact, say like PHE with surge drum on top.
Height of large vertical vessel may be an issue for pipework like wet return.
Thanks Ranger1, Yeah every plant is different but some cope with pressure fluctuations better than others. The cavitation start/stop control is a good idea. Asking for a rule of thumb was me being hopeful and in reality it's trial and error. Keeping the suction pressure stable in a large ammonia plant with variable loads is a challange. When loads reach set point is it better to leave the liquid out in the field by closing the suction valve or better to close the liquid and pump it out? If the liquid stays out in the field and warms up is that a greater load than opening a LLSV?
In some cases LLSV left on & throttles through BPV with Danfoss CVQ or ICM seen it in abattoir Beef chillers when down or holding temp set point. to temp.
If there is a load refrigerant will boil & need accumulator to call for liquid, but thinking load more stable than pump out of unit.
For ammonia, suction pressure should not drop faster than 1 psig per minute. I saw it in manual of pump manufacturer.Quote:
Originally Posted by RANGER1
Hi All,
Don't know about pressure drop limitation in LP vessel. But an allowable liquid variation = cooler volume + 1/3 wet return pipe volume gives us trouble free operation in almost all cases(dont know if this is acceptable for low charge system).
I prefer to oversize the pump suction to limit the velocity below 0.3m/s so even if there's bubble formation inside the vessel, it won't be carried inside to the pump.(IOR has a document about this)
One good benefit of the vertical vessel is that the vapor velocity remains the same irrespective of liquid level. In horizontal vessels, vapour velocity inside vessel changes with changing liquid level.
Make sure you pay attention to the NPSH of the pump at the pump flow you required and then sum about 1 meter plus of head collumn, depending also on the head losses you have on the pump suction side (valves, curves, filters, etc.)