Mixing of LPG with natural gas 2

[almostbroken]

I have a situation where we need to mix 20,000 bbl/d of LPG with 200 mmscfd of gas upstream of a compressor. The 6" LPG connection seems to be able to only "dump" the liquid into the gas line, leaving me with concerns about reaching steady state. HYSYS says that thermodynamically it will work. Any ideas on how to ensure it will?

 

[MortenA]

If I were you then I would route the LPG to the liquid separation train and let the LPG evaporate into the gas here. I know this will increase recycle and could cause the liquids to become cold - but i think that you will get problems "evaporating" the liquids into the gas just like that.

Maybe super heating the liquid under increased pressure and then use some sort of (packed) tower with a bottom stream going back to the LPG distillation tower?

Best regards

Morten

 

[Zoobie]

We do this...sort of...except that its downstream of the compressors and the injection rate is about 1/4 of yours (25 bbl/mmcf). The injection points go into the middle of the pipe and are equipped with spray nozzles. One of the constraints on the control system also is keeping the injection rate per unit of volume a healthy safety margin away from the dew point at process conditions.

Personally, I have always found mixing to be a tough thing to predict well so I always go to an expert. In this case you also have to deal with evaporation which makes it even more complicated.

Things I would look closely at: time to get from injection point to compressor, distance from injection point to compressor, routing of pipe (ie elbows, etc.), temperatures (and how stable they are).

Another area to look closely at is control. Is 20,000 bbl/d an average number? Does this fluctuate a lot (short bursts of high flow)? If so, you will probably have problems.

Besides nozzles, you might have some luck with others things to atomize the liquids (or at least get some small drops or films) and get good quick mixing). Static mixers or some sort of other piping internals may help.

 

[dcasto]

I'd save 700 Hp and headaches by pumping the propane into the hot discharge of the gas compressor.

 

[unclesyd]

I've seen a system similar to that posted by dcasto that I believe used a mixer by Komax. I don't have any details other than it involved rather large flows based on the pipe sizes.

http://www.komax.com/det-gasliquidcontacting.html

 

[MortenA]

A theoretical question:

If you had a lean gas and compres it to dense phase, and a liquid LPG stream, pump it to the same pressure, mix the two streams (assuming that the combined fluid remains in the dense phase - what would then happen? Would the two fluids just mix and for a uniform dense phase or what? Anybody know of experiments performed and perhaps observed visually?

Just curious.

Best regards

Morten

 

[dcasto]

How about this one. We had a pipeline full of ethylene at 900 psi, above its critical point. We wanted to push it out so we injected nitrogen at 900 psi at the end oppisite from the N2 injection we flared the ethylene and when the calculated volume of N2 made it to the flare, it indeed went out. Typical operation

One time we did this, we had to stop flareing for 4 hours. When we continued, the flare went out way sooner than it should. We set up the GC and watched the stream. It took about twice the volume to get the ethylene out becasue the nitrogen floated under the ethylene and stayed two phase.

You will need a mixer or have something get them to mix. but being supercritical is not any different.

 

[MortenA]

So as a minimum some sort of static mixer would be required?

Best regards

Morten