Help BEU type heat exchanger design

[knapee]

Dear Sir

It is desired to cool down 1260kg/hr of Mixed Gas having properties in below table from 70 C to 15C. This operation is performed with cooling water (56000kg/hr) from 10C to 20C. (Note tube side inlet p=3.075barg, delta P =0.075bar Shell side inlet p=3.7barg, delta P=0.7bar). The propose of this design is to use HTRI to meet the owner’s spec rule, and get the minimum heat transfer area or maximum heat transfer coefficient when delta P values of both fluid close to the maximum allowable delta P.
Unknown variable need to be key in in HTRI design mode:
1. Tube OD.(3/4 in for U tube)
2. Tube Wall thickness
3. Tube layout angle
4. Tube pass, length,
5. Baffle type
Unknown parameter could be calculated by HTRI design mode.
1. Tube count
2. Baffle cut, spacing
3. Shell ID.
(please see attache file)


Could you please give me a help about the following question?

1.Why the U type heat exchanger, tube pass can’t over 4 pass.

2.As I run into a lot of vibration warning message, I put vibration support inlet / outlet, but I still have acoustic vibration and bundle entrance / shell entrance velocity exceed 80% of critical velocity. Also, I already tried to reduce baffle spacing, increase the tube pitch, and increase nozzle diameters to reduce vibration, but in vain.

I know some acoustic resonance may be corrected by adding a deresonating baffle parallel to the cross flow direction to increase the shell acoustic frequency, but I don’t know where I can key in the HTRI.

Could you please give me any good suggestion? I really appreciate your comment and help.

Please Note.(Some explanation for U-type exchanger and other vibration terms)
Tube failures have been reported in nearly all locations within a heat exchanger. But those of primary concern are
1. Nozzle entrance and exit area. In these regions, impingement plates, small nozzle diameter , or large outer-tube limits can contribute to restricted entrance or exit areas. These restricted areas usually create high local velocities that can result in damaging flow-induced vibration.
2. U-tubes bends. Outer rows of U-bends have a lower natural frequency of vibration and therefore are more
susceptible to flow-induced vibration failures than the inner rows.
3. Tube sheet region. Unsupported tube spans adjacent to the tubesheets are frequently longer than those in
the baffled region of a heat exchanger owing to the space required to allocate inlet and outlet nozzles
. The higher unsupported span results in lower natural vibration frequencies of the tubes in this region. The possible high local velocities, in conjunction with the lower natural frequency, make this a region of primary concern in preventing damaging vibration.
4. Baffle window region. Tubes located in baffle windows have unsupported spans equal to multiples of
the baffle spacing. Long, unsupported tube spans result in a reduced natural frequency of vibration, and
tubes have a greater tendency to vibrate and suffer damage.

*Fluidelastic instability : Vibration owing to fluidelastic instability is avoided if the fluid velocity at each section of the heat exchanger is below a certain critical velocity. This critical velocity is a function of the tube geometry, its natural frequency of vibration, and the viscous damping effect of the shell-side fluid. If the fluid velocity at any location is higher than the critical velocity calculated for that location, potential tube damage can be expected.

*Vortex shedding vibration: the TEMA Standards include equations and graphs that allow calculation of the vibration frequency resulting from this mechanism. If the tube natural frequency is less than double the vortex shedding frequency, there is a potential risk of vibration damage. In this case, it is necessary to calculate the amplitude of the tube vibration, and this must not exceed 2 percent of the tube diameter.

Thank you so much.

 

[Montemayor]

I don’t consider this as a “large” exchanger – considering the gas flow and the cooling water flow (approx. 250 gpm). However, I don’t know the composition of the gas and the reasons for selecting such low pressure drops (approx. 1.0 psi on the CW tubeside) and relatively high pressure drop on the gas shell side (approx. 10 psi). The shell side is usually the lower pressure drop – as compared with the tube side.

Are you designing this heat exchanger or rating it? Before answering specific questions it is required to know what it is that you are doing – or expect to accomplish.

I don’t understand what you mean by: “1.Why the U type heat exchanger, tube pass can’t over 4 pass.” Could you please express yourself better? In the meantime, I will speculate that you mean that you don’t know why a U-tube bundle can’t be over 4 passes in the tube side. My response is that it can be as many tube passes as can mechanically be built-in. However, why do you need more than 2 tube passes?

As for your perceived tube vibration problems, I don’t understand your question / comments. Are you having trouble with the HTRI program’s vibration alarms? You state: “Tube failures have been reported in nearly all locations within a heat exchanger.” Is this a similar BEU unit that you are presently operating? Or is this someone else’s unit? Are these actual, personal tube failures that you are reporting or are you concerned and just quoting some literature or other reports?

If you are inputting the correct basic data in the proper manner into the HTRI program, I would not have any concerns. If HTRI gives you a warning, it is for a good reason – assuming that you are properly running the program. There are a lot of ways the tube vibration problem can be handled – some of these methods are in the process design and others are in the way the unit is mechanically designed and fabricated.

Since we don’t know your scope of work or your constraints on the design, we really can’t address the way or methods that can be used to resolve tube vibration problems – other than the manner employed in HTRI’s process design.

 

[TJOrlowski]

We've fabricated 20 pass U-tube exchangers here, 4 is certainly not the maximum number of tube passes in a U-tube exchanger.

Put your gas inlet in front of the U-bends (either above the tangent line for the u-bends, or better yet near the fixed head of the unit).

Add a blanking baffle to the rear end of the unit. This will add a lot of support to the tubes, and should eliminate your vibration issues. If it doesn't, try increasing the cut% of your supports, and removing rows of tubes below the inlet nozzle to increase the bundle entrance area.

You could also try using a type "J" shell, with your inlet centered, and two outlets. This will allow you to put a full support in the middle of the bundle.

You could also add thickness to the baffles.

You could also reduce the size of the tube holes in the baffles. In extreme cases we've even roller expanded the tubes to every baffle to combat vibration.

So many ways to skin a cat!

-TJ Orlowski

 

[knapee]

Dear Sir

Please forgive me because I have some typo in the last post.
1. The tube outlet P=3.075 barg, delta P=0.7bar(allowable), and the shell outlet = 3.7barg, delta P =0.075bar(allowable).
2. This is rating problem, but I don’t the number of shell and tube passes, diameter and number of tubes, clearances, sealing strips, tube pattern and pitch, tube length, tube length, type of baffles, and baffle cut, baffle spacing, and shell diameter, so I use HTRI design mode first to get a convergence result. ( In design mode, only the tube diameter, shell diameter, pitch is required user to input).
Then, HTRI calculated the other unknown variable, and I adopted this variable calculated by HTRI to make some change to get rid of the warning message.
3. I guess it is unreasonable to design the tube passes over 4 in the U type heat exchanger because it will make more vibration problem.

The last paragraph note is adopted from book “Heat transfer in process engineering by EDUARDO CAO”. And it is not HTRI vibration warring of my case. I am sorry I didn’t note it where this comes from.

My warring message is following

WARNING-First mode acoustic vibration is probable. A maximum Chen number of 14379.1 is calculated for the regions with a frequency ratio between 0.8 and 1.2. Consider adding deresonating baffles. Careful positioning of deresonating baffles eliminates noise.
NOTE-In acoustic vibration calculations, at least one frequency ratio for the first mode is greater than 1.2. For regions in which this condition occurs, the program calculates a maximum Chen number of 14379.1. Consider checking for higher mode acoustic vibration.
Bundle entrance velocity exceeds 80% of critical velocity, indicating that fluidelastic instability and flow-induced vibration damage are possible. Fluidelastic instability can lead to large amplitude vibration and tube damage.

Shell entrance velocity exceeds 80% of critical velocity, indicating that fluidelastic instability and flow-induced vibration damage are possible. Fluidelastic instability can lead to large amplitude vibration and tube damage.

Shell exit velocity exceeds 80% of critical velocity, indicating that fluidelastic instability and flow-induced vibration damage are possible. Fluidelastic instability can lead to large amplitude vibration and tube damage.

Xvib can be used for a more detailed analysis of individual tubes in the bundle.
Thank you so much for your help and I appreciate you to give me this valuable comment.

 

[knapee]

Dear Sir

I have to correct my last post again, and very sorry for this kind careless.

The shell outlet P should be 0.1 barg instead of 3barg. The shell inlet/oulet nozzle is 400mm class 150 WNRF, tube inlet / outlet nozzle is 100mm class 150 WNRF.
The tube layout (please see attachment)

The problem is that I don’t know where is a good location to put dersonating baffles in the tube layout to eliminate noise.

And, another small trivial issue is that I don't know what is WNRF meaning. Could you please kindly give me any comment?

Thank you so much for your generous share of knowledge and help.

 

[knapee]

Dear Sir

Someone suggest me that I should change the tube layout angle from 30 degree to 60 degree to make the vibration warning disappear. And, indeed it really happened.

But, I just don’t get it why the rotated triangular (60) and triangular(30) will make so much different, and it seems that the tube no is much less than 30 case in the result.(Please see the drawing in the attachment)

And, it also confused me, from my limited experience, it seems that the rotated triangular is not often used cause it has no heat transfer or pressure drop advantage over triangular. I don't know why 30 degree tube layout seems better than the 60 degree tube layout in this case.

Any comment and suggestion will be highly appreciated.

Thank you so much.

 

[QPAini]

Dear knapee.

WNRF is W= Weld
N= Neck.
R= Race.
F= Face.

 

[Montemayor]

The "R" in WN RF stands for "Raised", NOT Race.

 

[srfish]

The 30 degree triangular tube pitch is better because it has lower vortex shedding frequency than 60 degree triangular tube pitch.