sshep
Chemical
- Feb 3, 2003
- 761
Our ethylene plant uses a three stage expander on the H2 and methane demethanizer product vapor to recover the refrigeration power for the demethanizer reflux and feed chilling. The liquid methane and hydrogen/methane vapor from the expander outlet are the only refrigeration available for the plant below the -150F level. Recompressors on the other side of the expander shafts bring the hydrogen up to 120psig. It is the recompressor discharge pressure which is controlled, the expander outlet pressure is the result from the power balance. Currently our expander speeds are lower and discharge pressure higher than design, with the result that nowhere near design methane liquid is being generated and the refrigeration being produced is low. This is leading to excess ethylene losses. It is believed the expanders worked at design shortly after being installed, but these problems appeared after a subsequent plant restart.
My question is:
Since the expander discharge pressure is not controlled, can there be more than one stable operating point and/or feasible start-up path issue for this type system?
My thought is that the additional methane that must be compressed is creating added power requirement for recompression that will not exist when the process actually gets to design isentropic efficiency on the expanders. I want to bleed some gas to fuel (or flare) before the recompressors to get the expander speed up to design, but there is no piping to do this. Tie-ins do exist and I think the idea has a good chance of success, but I have met some resistance because the system seemed to be working at one time. Any input is appreciated.
My question is:
Since the expander discharge pressure is not controlled, can there be more than one stable operating point and/or feasible start-up path issue for this type system?
My thought is that the additional methane that must be compressed is creating added power requirement for recompression that will not exist when the process actually gets to design isentropic efficiency on the expanders. I want to bleed some gas to fuel (or flare) before the recompressors to get the expander speed up to design, but there is no piping to do this. Tie-ins do exist and I think the idea has a good chance of success, but I have met some resistance because the system seemed to be working at one time. Any input is appreciated.
