4 replies to this topic

[Sadananda Konchady]

Cooling water suction pipeline layout guidelines I have come across suggest at least 3 times pipe diameter length before the pump suction flange.

Higher suction line velocity would seem to reduce the chance of suspended solids building up in the suction line according to one reference. Could the suction line be the same size as the discharge line (which could be sized between 2 to 2.5 m/s velocity depending on the distance of the cooling water discharge outlets in the plant)?

When there is a standby pump would it be preferable to use a suction line header or provide two independent straight suction lines to the pumps? The second option seems to be a more economical choice since it avoids the need for a large header.

[Art Montemayor]

Here are some comments, based on my field experience:

I have used a distance of 5 times the nominal suction pipe diameter as the length of the CWT suction line to the circulation pumps. A suction line size of 10”, for example, is common for small towers and if you used 30” as the length, you wouldn’t have enough space for maintenance – or even your flanges! The longer the “straightening” distance (within reason), the less the entrance losses and turbulence as the water enters the impeller’s eye. Be careful how you define the suction piping length. Look at the sketch I’ve furnished to get my strict definition of what I’m talking about as to what constitutes the “straight", true suction pipe run length prior to entering the impeller eye.

You should not be talking (or considering) about designing for suspended solids when you design a cooling tower. If you really have suspended solids, then you have a more serious problem than a suction line into the circulation pump. Besides, any solids – like wind blown sand - will precipitate to the bottom of the tower sump and stay there. They should not be pumped out. This is ensured by locating your cool water pump suction nozzle at least 1 foot above the tower sump’s bottom. This is normal practice. The lower the velocity of the suction fluid entering the pump, the better the operation – and the closer you can locate your pumps to the sump. Extrapolating this effect to the extreme, you can wind up with submersible pumps inside the sump – which I have also used.

Standby pumps should be installed independent of each other – with their own suction line connected directly to the sump. There is no advantage or benefit from installing an external header to the CWT sump.

Art Montemayor
 CWT_Pump.xls   19KB   273 downloads

[joerd]

One of the things to bear in mind is to keep a good margin between calculated available NPSH and the required NPSH of the pump. Due to the dissolved gases in water from a cooling water tower, and the low elevation difference between liquid level and pump, vibration/cavitation can easily occur.

[Art Montemayor]

I guess Profe's reference proves that I didn't just pull the 5 x D figure out of my hat. I knew I had that figure from a prior experience and that it was backed up by similar experiences. I just didn't remember the other sources that also acknowledge the figure.

This is a very good recommendation by Profe. His illustrations give a good example of what can be found at the ITT website. ITT and Goulds Pumps have merged and the excellent Goulds Pump Fundaments information can be found there at:

http://www.gouldspum...t_Technews.html

and also at

http://www.gouldspum...at_pf_0002.html

I also recommend every engineer to visit this site (especially the younger set and those aspiring students who want to get on the learning curve faster than others) and download all the important pump information available there. If you highlight and paste, you can form a 200-page Word for Windows pump manual with the information found there. I still have a hard copy of the Goulds GPM pump manual from which a lot of this information came from. That hardcopy - like all manufacturers' catalogs - is no longer available I believe and that is the reason I would strongly advise all young engineers to spend approximately 3 hours downloading this information. You have to have it handy and in a much better organized document. The website lacks organization and is difficult to find the information you know is there.

Good job, Profe.