6 replies to this topic

[GS81Process]

We are planning to replace a tube bundle on a shell and tube heat exchanger at my plantsite. I am looking to use this opportunity to make a few design improvements to the exchanger. The basic process conditions are going to remain the same afterward, i.e this isn't part of a plant revamp.

The relevant information from the existing exchanger datasheet are as follows:

Exchanger Type- BEU, Horizontal
Heating Surface- 26.74 m2.
Duty- 473 kW
LMTD- 32 C
Clean Transfer Rate- 763 W/m2C, Service- 436 W/m2C
Shell ID- 406 mm
Tubes- 70 U-tubes, 19mm OD (11 BWG) x 3800 mm length, tube pitch 25.4mm, square rotated.
Baffles- single segmented, 29 with center spacing of 132 mm, 19 dia.% cut
nozzles- shell inlet 101.6 mm, shell outlet 101.6 mm, tube inlet 76.2mm, tube outlet 76.2 mm. Shell side inlet at top, rear head. Tubeside inlet at top.


Shell side
60wt.% ethylene glycol, 40wt.% water (liquid)
72,000 kg/hr, max. inlet temperature 25C, 120 kPag.
Fouling Factor- 0.0004 m2C/W
Allowable pressure drop- 35 kPa
Design temp- 60C, Design Pressure- 6 barg


Tube Side
Ethylene Gas
8040 kg/hr
max inlet temperature 95C to be cooled to 43C.
Inlet pressure 11,700 kPag
Allowable pressure drop- 35 kPa
Fouling Factor- 0.0004 m2C/W
Design Temp- 160C, Design Pressure- 184 barg


Based on some heat transfer analysis, my proposed modifications are:

- Increase tubepasses from 2 to 4. I expect that I will have to reduce u-tubes to 58 because of the space taken up by the pass-partition lanes.
- Replace single segmental baffles with double segmental to reduce shellside pressure drop.
- Increase baffle spacing from 132 mm to 158 mm, and %cut from 19 to 25.
- Add 2 rows of impingement rods to top of bundle to protect against erosion based on TEMA requirements.


I would appreciate any comments on the above proposal.

Has anyone done a similar project? What mechanical design considerations I should be aware of?

Edited by GS81Process, 29 December 2011 - 03:16 PM.

[DB Shah]

You should be more worried about tube side DP. As you are increasing tube pass, tube side velocity and pressure drop will go up. With very rough estimates, your present DP on tube bundle is ~ 0.03 bar & 2 m/s velocity, with 58 U tubes and 4 pass you may have additional 0.1 bar DP and vel of ~ 8 m/s pl check this data in a simulator

I do not think you should need baffle changes. When you are reducing number of tube count (from 70 U to 58U tubes) why should shell side pressure drop increase, it should reduce.
You shell side nozzle is operating approx > 2 m/s velocity, check rhov2, may be instead of impingement rows, you might be OK with increased shell nozzle size from 4" to 6"

[GS81Process]

Thank you DB Shah.

I checked the tube side pressure drop and velocity and found them to be acceptable with the new proposed configuration.

The shellside pressure drop is too high right now. I agree that reducing tube count to accommodate 4 tubepasses would reduce shell side pressure drop somewhat. But most of the reduction in shellside pressure drop would come from switching the single segmental to double segmental baffles, because it will reduce the shellside velocity by roughly half and thus the pressure drop by roughly a factor of 4.

I agree with your comment regarding shell nozzle size increase. If this was a new exchanger I would increase the shell nozzle size and inlet glycol piping size. Unfortunately the inlet piping and nozzle change is not an easy job since this exchanger is currently located in a fairly tight space. So I am attempting to resolve this issue by adding impingement rods and an inlet tube support below the shell entrance nozzle to reduce vibrations.

Edited by GS81Process, 30 December 2011 - 09:05 AM.

[DB Shah]

GS81process,
I appreciate your concern as I too am working in a operating plant, You intend to go for 4 tube pass to increase Ud. If I summarize your modification-
New tube bundle
New tubeseet
New channel cover
New baffles.
All (except baffles) are high pressure elements. Only thing you dont change is low pressure shell. Hence replacing the shell will be a small addition in the cost.
Considering above, you can think of replacing the exchanger with same length with a better design.
One more point, if you change no of passes, the tube nozzles location will vary, present high pressure line may need modification and stress analysis. It will be a better option to maintain two tube passes, redesign hex with same tube length. If shell dia varies you will have a very small job (cutting & welding) without any other change in high press line. We have done this in one of our Zr exchanger. The constraints - short shut down time, no major jobs on high press lines.

[srfish]

GS81Process

I believe you have made all the right proposed changes to the existing heat exchanger. Since the tube pass change will give a big increase in the tube side heat transfer coefficient and it is the controlling side, the reduction in the shell side coefficient will have a minor effect.

You may not need to have a new channel but only change the pass plates.

[DB Shah]

dear srfish,
yes the channel will remain same.
With so many changes in high pressure side, will not a new exchanger with better design a good option. Another thing is that present tube pitch is 45, maybe this was due to fouling/vibration point of view, with new bundle design triangle pitch can also can be explored. I will be happy to have your experienced views, as we encounter such modifications often in our plant.

[GS81Process]

DB Shah,

I believe the square tube layout was selected because it allows for cleaning between the tubes in this fouling service. The rotated square (45degree) tube pitch was likely selected over the square profile (90 degree) because of better heat transfer.