Slug flow vs Wave Flow

[gte682n]

Hello,

I am evaluating the flow regime of a 3" line that handles a 2 phase mixture. Aspen Plus predicts slug flow in the 3" line when I reach my destination pressure. The vapor volume fraction is 87% and the mixture velocity is 48 ft/sec at the end of the line. When I increase this line to 6", I am in the wave flow regime for the entire run, and see a max velocity of 11 ft/sec at the end.

There total pipe run is ~300 ft. I would recommend to replace all of this 3" pipe with 6" to remove the possibility of slug flow.

This fluid becomes two phase after exiting a feed pre heat train prior to entering a vacuum distillation tower. Installing a back pressure valve is not an option as that would bring the feed pre heat train very close to relief valve setpoints.

Are there any other options that I should consider, and is Aspen trust worthy for these flow regime predictions?

Thank you.

 

[LittleInch]

Yo could try a short section of 6" line at the end point to see if it smooths the slugs out.

Or add a vertical loop just before the end - that can act as a "slug breaker".

However your velocity in the 33 lie is at the high end for two phase flow.

Aspen is pretty good I think , but designation of flow regime is not an accurate assessment in any software.

does slug flow cause an issue? Can you introduce a break out tank.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

A crude V/ L phase break separator made out of piping upstream of the feed nozzle on the vac tower would be a good idea. The flow regime should be checked for turndown also , which may be 25-30% of design flow.

 

[TheChemengers]

What is the slug length Aspen has calculated? Are there significant straight lengths where slug flow could fully develop? Going from 3" to 6" is a significant jump, both in cost and space terms. I wouldn't be making that change without a thorough analysis.

 

[gte682n]

Hello, sorry for my delayed response.

LittleInch, I am not familiar with a vertical loop slug breaker - how would this work?. I have searched the forum and see "slug catcher" mentioned. Are you referring to the same thing? There is visible movement in the line when it approaches the tower. The pipe routing is also not ideal in this area since there are many elbows and changes in elevation, both up and down.

TheChemengers, I am not seeing a calculated slug length in Aspen. The only output I see with the flow regime is for the inlet and outlet of my pipe. I have changed the length of this pipe to see where Aspen begins to predict slug flow, and is it very early in the pipe - the first 20 ft of a 300 ft line.

The actual pipe routing is straight for the most part, until the final 50 ft or so. Since there is no visible movement in the line in the straight run (even though Aspen predicts slug flow here), I am thinking of recommending increasing this last 50ft of pipe to 6" where slugs are most noticeable to make it more economical.

Thoughts?

Thank you for your responses.

 

[LittleInch]

The last 30ft at 6" would probably be enough.

No what I mean and what I think george also means is a vertical loop up, along and then down.

What tends to happen is the liquid builds up in the horizontal section, slugs up the vetical section / riser and then as it falls away down the far side, it naturally breaks up the slug into a more uniform manner.

However I suspect at 3", you're going to get slug flow even in a short section.

Can you sketch isometrically the last 50 feet?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[gte682n]

Here is an iso of the line in question.

Thank you

 

[georgeverghese]

A slug breaker relies on another LCV on the liquid outlet line of the slug sep in order to smoothen out slugs going to the downstream tower - which isnt going to help here since you've got design pressure limits on the preheat HXs'.

Your sketch indicates a high point section (within the cloud) - this will mean that liquids will back up all along the upstream piping and this may be the major reason for this slugging - you'll probably see that on the Aspen output which should show a high liquid holdup in the piping section upstream of the high point. This liquid holdup will be even higher at turndown.

Can you find a more direct route to the vac tower that avoids high points ? You'd then have a fairly consistent holdup for the entire piping route.

 

[TheChemengers]

Looking at the iso, it would be good to know the slug length. Obviously no lengths on the iso, but im guessing that section you've clouded is comparatively short? You need a good straight length of line for a slug to form. if the calculated slug length is comparable to the lengths of straight line in that section then its possible to infer that fully developed slug flow will not occur - although a bit of engineering judgement is required and of course err on the conservative side.

Another question is what is the slug velocity? you could get some stress calculations completed to determine if in fact the slugs would be a concern. If the velocity is low enough it may mean the slugs will cause no damage. I know you're using aspen so perhaps it does provide this. I've mostly used HYSYS and it provides slug length and velocity.

Certainly I wouldnt be recommending moving from a 3" to 6" line purely on the basis of an aspen result - id be wanting a more thorough evaulation. But i think you're doing the right thing so far in thinking through all the various options.

 

[georgeverghese]

Another suggestion to a complete reroute of this piping would be to add a new loop line, teeing out from where you've got the cloud intersecting the piping, to tee in ( without any high or low points in this loop line) at the last straight length section which is just before the feed nozzle to the tower. Only gas would then be forced to go through the current riser section , while some gas and all the liquid would go through this new loop line. This will most likely reduce liquid holdup backing up in this entire piping run and reduce slugging in these vertical risers.

 

[LittleInch]

That's a complex bit of piping so I'm not surprised it's shaking a bit with liquid slugs coming through at > 10m/sec....

Bigger pipe should fix it.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.