9 replies to this topic

[afd]

We have 2 centrifugal pumps with double suction. One is running and one is standby.
when we change from pump A to B it cavitates. liquid being pumped is saturated water at 7 barg. flow 400tons/hr.

To overcome this phenomenon we raised the feed vessel level to maximum and still the phenomenon existed. So we throttled the running pump discharge to reduce the flow to 250tons and then started the standby and it ran well.

My explanation is the running pump is sucking all the suction line liquid and therefore when the standby is started it is starved and cavitates.

what can the other measures be to prevent this occurance. As an operator I do not want to reduce the flow of the running pump.
The standby pump is not having a suctions strainer and the suction piping is identical. This is a recently observed phenomenon.

Afd.

[Art Montemayor]

Hello, Afd.

It’s good to hear from you again.

What you describe is exactly what can happen if you have two centrifugals – one on standby service – both connected to a common suction manifold. Pumping saturated water is yet another hurdle to overcome and should be seen as a special application where accurate, conservative design is a priority. Obviously, the NPSHa for the standby pump was not done correctly.

The suction pipe size you are dealing with is probably a 16” (or hopefully 20” – since 18” is a “bastard” size). This size of pipe means you, as an operator do not have many workable options available to you since that size of pipe takes up a lot of real estate and is a costly modification.

Not knowing more than what you’ve told us, I have to guess that the existing common suction manifold is a futile attempt to try to operate both pumps during a change-over. I would seriously look at a separate, independent suction line to the stand-by unit (or to the regular, on-line unit, if that is more reasonable). I believe you are correct in guessing that the common suction line is either undersized or configured in such a way that the suction flow favors only one of the pumps. This error cannot be remedied in the situation you presently find yourself. You really only have two choices:

1. Do a total redesign and reinstallation of a new, larger, common suction manifold; or,
2. Install a separate and independent suction line (again, approx. a 10”) to the stand-by unit.

Depending on the NPSHa and the cleanliness of the water and the system, you may not have to install a suction strainer. I normally would expect that it is easier and more feasible to install a separate and independent line while you are running one pump than it would be to re-install a larger and re-designed suction manifold for both pumps.

I hope this helps you out. I like your type of pumps. Horizontal, double suction centrifugals with outboard bearings are a good, solid, dependable design. This type of pump should be well cared-for because it can give many years of faithful, steady performance - if installed and treated correctly. Good luck.

[toronto engineer]

I still don't understand why this phenomoena should occur when you have 2 identical piping on the suction. why should one pump be starved over the other when both are identically positioned? Could it be that the standby pump, though similar has become different and has a higher NPSHr than the normally operating pump?

[pawan]

Though Art has already defined your problem & explained it I would like to add following.

1. In your case, fluid is at saturation condition so any change in velocity (which is getting doubled in this case when U run pump A while B is already running) will lead to drastic drop in NPSHa.
   
2. Also, over a period of time the performance of A might have altered as you rightly pointed out

For solving it the best & fastest is try to slightly increase the pressure of suction vessel if possible. U can easily estimate additional line drop when velocity goes high. It is possible if you require this only during changeovers & not for increase the capacity by double.

Provide some sparkling or trickling of CW on the suction flange of the pump/line so that some margin is available by reducing the temp by 1-2°C.

Two other options are already mentioned by Art.

[Technocrat]

AFD,

Please tell us which pump cavitates, A or B?
Please send us the suction piping isometric.
Please tell us how do you take the changeover, we need to check the procedure.

Regards.

[mishra.anand72@gmail.com]

In given situation, your pump change-over procedure is just good. Going for larger suction manifold is possible during annual shutdown. Going for separate and independent line is possible in running operation and seems to be good solution.

[Padmakar Katre]

Dear AFD,
Please could you answer my following points,As per my information double suction pumps can work at a 27% lesser NPSH. In case of symmetrical suction piping there should not be problem of cavitation provided the design is correct. The following points come in my mind, could you please reply thses points ,
1.Could you tell me when this problem is started now or is it from the beginning. When did you removed suction strainer of the Pump B.
2. As it is with the single pump (specifically with B pump) then only thing to check is the suction valve status (is it in a good condition) as already you mentioned that the suction mainifold is symmetric. So Suction Valve is only thing which I can give emphasis on.

[Rama]

This might sound silly, but my 40 years in a chemical industry has made me over cautious:
Are you sure that both the pumps are absolutely Identical? Please check if the impellers are of the same diameter. There are times a slight reduction in the diameter is not noticed, but can cause havoc in limiting situations such as yours. Also please check for any blockages in the system - valves, pump suction etc., since you say that you have no strainer in the second pump (.

All the best.

Ram.

[Technocrat]

AFD,

What is the reason you removed suction strainer of B? By removing it you have reduced the resistance in suction line to B and thus B is sucking more liquid which may be causing less suction to A and cavitating A.

Regards.

[afd]

Sorry for the delay in getting back

The pumps are running for a decade now. I think the problem has surfaced off late this year. When we change pumps we need to throttle the discharge of the running pump considerably and then put the standby in service.

What is the mechanical aspect that can cause the pump not to take suction?

The pipeline is identical.

Afd