Lube oil carry over

[B62plant]

I export gas via wakashau compressors which need lube oil injection. I have a filter coalescer fitted to take out lube oil carryover but I am only removing about 90% of the liquid. Can anyone recommend a better method or better system than the PALL unit I am currently using. The client stipulates zero liquid in sales gas.

 

[georgeverghese]

Does the lube oil at the compressor cylinders come into contact with O2 / moisture laden air anywhere in this lube oil circuit? If so, perhaps contact with O2 and water vapor are accelerating degradation leading to formation of light ends ?

Carbon bed adsorbers can remove lube oil laden in gas streams, but at this pressure, it is very likely co adsorption of heavier hydrocarbons in the gas stream will dominate, leading to premature saturation of the bed.

So departure conditions are 1200psig/ 88degF; what is the normal arrival pressure and temp of this gas at the other end ?

 

[B62plant]

No, it's a hydrocarbon system there is no air/O2. The pressure and temp are consistent and similar at the end of the 16km pipeline other than usual frictional losses. After 16KM it enters the client network and that's where they are getting the 20-30 litres of lube oil out. We are the only importers at this point so I know it's coming from us. Only other thing worth mentioning is that I am seeing black solids in the filters (when changed out)had them analysed and it is iron, carbon and silica based. Probably corrosion of the piping.Agreed, carbon bed would be easily saturated and create too much delta P.

 

[B62plant]

I'm having iso kinetic testing done by slipstreaming through a portable pilot plant. This will show me exactly what particles are there with a laser particle counter. $30,000 a day this is costing !!!

 

[georgeverghese]

You may be able to tease out a refrigerant stream from within the gas plant with the compressors running on full speed with some recycle, which you could use to chill this LO laden gas stream at a compact, high eff HX and then route it back to the filters, but I doubt the refrigerant duty would be adequate for 90mmscfd to get it back to the contract dewpoint value. Some thought required here.
If the PECO filters work ok, you're in luck, else I guess you can see the writing on the wall:- replacement of these LO injected recips for a battery of oil free recips or a single oil free centrifugal. Enabling the additional refrigerant duty may turn out to be cheaper.

 

[B62plant]

I think I'd be creating more problems than I have already trying to pre chill the gas. I'll see what PEKO tells me and post results.

 

[georgeverghese]

BTW, Shell have a proprietory design technology for clean service high eff vap / liquid separation called the SMSM ( schoepentoeter - mistmat - cyclonic swirldeck - mistmat) which is made available to the general public through Shell Global Solutions (SGS). Separation eff of >99 % of the influent liquid stream is claimed. We have used this many times in high pressure (100bar and more) applications. Capacity turdown is 10:1 if I remember correctly. If you need higher turndown, I'd say use 2x50% capacity SMSM separators strung up in parallel to give you 20:1 turndown. Pressure drop would be less than 1bar. If this is of interest, maybe interpose this SMSM between the compressors and the existing filter-coalescers.

 

[B62plant]

I'll take a look thanks but our process team think the gas oil content is too low for a cyclone system to work