Hi expert

I am working on a forced circulation crystallizer circuit connected to pump & from pump to heat exchanger & heat exchanger back to crystallizer.(refer file attached)

I am having the below queries:

1. As the routing is very stiff, i need to include bellows in each line.
Can somebody tell me the preferable location & type of bellow for these type of typical circuits.
2. As this is a slurry line with high flow rates, is there any rule of thumbs that need to be considered for this typical circuit.

Expert opinion is welcome.

Hi Vick,

I would approach this first without considering bellow. Check the code stress, pump nozzle loads, heat exchanger nozzle loads, crystallizer nozzle loads, flange evaluation check, local stress on support check (if applicable).

From there, you can decide if there is really a need to put bellows and where to put it if space or hydraulic is a really constraint to re-route the line. You can take advantage of using advanced tools such as FEA Tools for the flexibility & K factors (say, for tees, elbows, exchanger/nozzle junction), as sometimes it might help (especially for high D/t ratios).

I would also take a closer look on the thermal profile of both the line & equipment (as thermal gradient may come into play) by communicating it to process engineer.

Any other opinion is highly appreciated. Just correct if any wrong statement I've made.

Cheers!!!

Hi Vick,

If you have a concern regarding high flow rates, you can consider referring to EIG for screening if FIV will be a concern.

Any other opinion is highly appreciated. Just correct if any wrong statement I've made.

Cheers!!!

Thanks Borzki for your response.

However, as per stress analysis, this circuit is not working without bellows.

Therefore, i am going with attached arrangements.

Any opinion/Suggestion on this is highly appreciated especially for type & location of bellow considering slurry line with high flow rates

Thanks

Hi Vick,

As long as you satisfy all the criteria in your project specification you should be fine. Please note that expansion bellows have a certain criteria such as unity check. Refer to EJMA. What's the temperature of the line and the density*velocity^2? The density*velocity^2 will give you a way to gauge if FIV will be a concern. The temperature of the line will tell you whether you are operating in the creep range and will give you an idea if elastic follow-up will be a concern.

I'm not so expert in bellows because in my twenty two years in the industry, I only experience it maybe 4 to 5 times. Tie-rod is important aspect in bellows to absorb pressure thrust. Check also if your system will experience high number of cycles, since bellow is a fatigue sensitive component make sure you capture the cycles in datasheet and life span.

Any other expert opinion is highly appreciated. Just correct any wrong statement I've made.

You can try to discuss with lay-out people if they can give some space in the plot plan to have a flexible piping.

Cheers!!

Vick,

Consider mounting the pump on a spring base. See Goulds Pumps Manual GPM 9 page 267. The photograph is of a Goulds Pump/motor mounted on a spring base and looks very much like one of my installations for a circulating pump on a strong black liquor circuit in a pulp mill. The pump is NOT bolted to the base. It is free to "slide around" but it will not "slide around" excessively because the pump suction and discharge are bolted to some fairly rigid piping. The only movement will be due to thermal expansion of the piping connected to the pump suction and discharge. Again, the pump is free to move around to accommodate the thermal growth of the piping.

Pat LaPointe
patrick.lapointe@pcllengineeringltd.com

Thanks Borzki & Pat for your response.

However, considering pump with spring is always risky option. we need to model basplate/motor etc properly in CAESAR II simulation.

Instead i thought to put the exchanger on spring.

At present, i had considered PTFE hinged bellow & PTFE lateral bellow in analysis.

Whether PTFE bellows are ok for this type of slurry with high flow rates or we need to consider some other options (rubber or metallic with liners etc) ?

thanks

Hi Vick,

You can consult to expansion joint vendor with suitable experience in slurry piping. So far I haven't experience such application. Ultimately, the vendor needs to consider the fluid in the expansion joint in the serviceability of the EJ.

Cheers!!

I would use the EJ on the pump discharge as is indicated on the sketch.
I would also leave the exchanger on solid support.
You may also see that on internet there are a lot of pictures of your layout and you may see that sometimes other two belows are used on the verical spool of the pump suction and the other between the exchanger and the crystallizer.
It depend a lot on the temperatures involved.

Thanks Danb for your response.

Pump discharge EJ should be untied one per sketch.However in reality it is not working due to pressure thrust (ope pressure: 3.5 barg, pipe size: 24") of EJ on pump discharge and exchanger inlet.

Similarly the EJ on vertical spool for pump suction is not working as for vertical configuration again only untied EJ is only option that is creating problem due to pressure thrust (ope pressure:2.7 barg, pipe size:24").

Hope you understand & accordingly we are going with the configuration of caesar II snapshot in above.

Any idea further from your side can help.

Maybe you can move the pump. You have only a sketch, doesn't necessary have to be exact. The sketch shows a tangential nozzle so the exchanger is not alligned with the vessel.
Some slurry pumps are special design to withstand the thrust but you must talk with the vendor.
Slurry service is different than normal service.
If you will use rubber belows, you will have to use filled arch belows which will increase the stiffness.

Sometimes the pump suction is TOP (I know is somehow strange)

Whether or not a given expansion joint is suitable for slurry service is for the manufacturer to tell you. I would be disinclined to use them as I expect any solids that can get trapped in them ruins their performance. Whether or not there are rubber bellows that can be used instead, I can't say. The pressure isn't too unlike industrial tires for vehicles, but I can't tell you what the life expectancy of rubber in your service will be, especially if you're saying it's high velocity.

With regards for springs on vessels, I would recommend a write-up on how these springs should be treated for and after installation. Something along the lines of:

"Springs should not be released until after piping is installed."
"Springs should be brought back to their set points and locked prior to breaking of flanges."

Vessels on springs have historically been a pain to deal with because it's not documented what the analyst expects them to be used for.

I would anticipate that you could eliminate the vessel springs altogether if you have a sufficiently flexible expansion joint between the vessel and pump top nozzle. If so, I would anticipate you'd support the vessels at the same elevation if possible to avoid nozzle stresses, even if it means holding the smaller vessel up by its nozzle. But this is completely dependent on design temperatures and weights. It's generally worthwhile to beef up one nozzle to handle a bit worse sustained loads and no thermal loads than it is to beef up two nozzles, add 4 springs, expansion joint between equipment, etc.

But if they already ordered the equipment, then you might already be backed into that corner.

Hi Dave/Richard/Caesar II Experts,

Can you please check whether the pressure thrust force & its direction in below snap is correct or not?

Hi Vick,

Is the bellow untied? If untied, the pressure thrust will be felt by the elbow once the pipe is pressurized. You can try blowing a piece of straw. But I have second thoughts in using untied bellows.

Just correct any misconception fellow stressers!!

Cheers!!