Cooling Tower Cell Arrangement Design 1

[aquadigio]

Dear All,

I am currently involved in a mega petrochemical complex project. The overall cooling water consumption is more than 130,000 m3/h (10& margin) and the initial design is to split to users into 2 main clusters, so end up with 2 block/cooling tower in two different locations within the complex.

There are 18 and 16 cells respectively for each cooling tower (one basin each) with 4000 m3/h per cell.

The issue is we are at the bottleneck on the cell arrangement. There are two options:
a) 9 x 2 (rows) and 8 x 2 (rows)
b) 6 x 3 (rows) and 5 x 3 (rows)

The above 2 options are being discussed between ourselves (project owner) and contractor. Contractor is preferring option 2 with the reason of easy maintenance and piping/nozzle layout
However, project owner side is in the opinion that by having three rows will not have maintenance issue.
By having 2 rows of cell arrangement(option a), the overall unit length will be too long compare to option b.

Appreciate your feedback. Thank you.

 

[ash9144]

Assuming you have countercurrent CTs with air in upflow direction you can't use option B. You will have cells which are surrounded by cells. Which could only draw air from bottom which would be drawn across your basin and saturated thus having no cooling effect.

 

[aquadigio]

Thanks ash9144, for your comment.

Also, just wondering the option B will have the STARVING TOWER FOR AIR scenario since there is a row of cell in the middle of the entire unit apart from what you described earlier.

In terms of maintenance and piping/nozzle layout consideration, what is your opinion on this aspect for both options?

Thanks in advance.

 

[GHartmann]

You need to have some concept of how you want to feed the towers and where the pumps will be installed.

For the two parallel units there could be a common header for CW return between the 2 units, with pumps located on the outboard side.

For the 3 units in parallel, I think (if tower lenghts are E-W) you would have the headers all on the N or S side of each row, with pumps on the opposite sides.

Here is one that was installed 1056 ft long (100,750 m3/h):

http://www.geapowercooling.com/opencms/opencms/gpc/en/case_studies/forney_case.html

I suggest you involve the aid of some good piping / layout designers / engineers to get a better handle on the proposed installation.

One thing for sure you need to have the cooling tower vendor tell you the minimum spacing between rows.

 

[GHartmann]

Here are two rows of induced counterflow cells that I would think are too close together.

http://www.classikcoolingtowers.com/modular-cooling-towers.html

Note that the risers are both fed from the middle "aisle".

Good luck!

 

[aquadigio]

Thanks Ghartman on the info provided. However, the suggestion by you on the 3 units in parallel are separated whilst our design is to group all 3 units parallel together (back to back).

Anyhow, it seems like option B with all cells grouped together would probably need further enhancement.

 

[GHartmann]

The cooling tower vendor will need to change the total area for the 3 x 6 cells vs the 2 x 9 cells as the center cells will not get as much air flow if they had open sides. I assume the units are the induced draft counterflow design. (Fans on top pull air through the tower cells).

Looking on the web I only see a few 2 row back / to back installations and no 3 row back to back to back units.

I suppose that is just a way too keep the fan blades from getting too long in radius. So maybe my first statement is incorrect.

This is something I would put heavily in the vendor's area of expertise.

 

[aquadigio]

Thanks for the feedback, will try discuss with the team. Cheers.

 

[mjohnqa]

just two points: 1)it is better to use 2 rows only as this not only enables ample air to circulate it also saves a lot in maintenance, as it may required heavier cranes to reach on top near to the middle row.
2) no. of cells can be optimized by using more blades per fan and hence the total operation and maintenance cost can be saved.

 

[Longinthetooth]

Echoing all the previous posters. You must consult the cooling tower vendor and adhere to his minimum spacing requirements, otherwise any process guarentees are off.

 

[aquadigio]

I believe 2 rows (option A) is a better design, however the project owner is yet to response on all the facts that we have given, perhaps they are thinking a way to 'fight' back to option B, hahaha...

Regardless, thanks you guys for all the valuable inputs...Cheers!

 

[GHartmann]

Your cooling tower arrangement is an extremely large application. For towers this large many people will choose the large parabolic natural draft towers.

http://www.sciencephoto.com/media/342223/enlarge

I would want to choose a supplier / field erection contractor that has demonstrated prior experience in similar projects.

This may put some of the smaller suppliers out of the picture.

Again, I think the decision should be governed by what the tower supplier recommends.

 

[aquadigio]

Hi Ghartman, we did have the same idea. However, it was turned down due to that the CAPEX is way too high compare to typical cooling tower.