9 replies to this topic

[SAprocessGD]

Hello,

My topic is regarding pumps and negative suction pressure. In my system I am designing, liquid is drawn from an atmospheric storgae tank and flows to a pump. The friction losses are quite large on the suction piping which will result in a negative suction pressure (i.e. -5.0 kPag) at the pump, resulting in pump lift.

My NPSHa is adequate (it is above 3 ft) and it is positive so I am not concerned about my net positive suction head available and any cavitation.

What I would like to know is: Can you design a centrifugal pump which has a negative suction pressure? If so, what does this negative number signify, what is occuring at the suction of the pump?

If I missed anything that could help you answer my question, please let me know. I really do appriciate any comments on my post. Thank-you!

[ankur2061]

SAprocessGD,

Check out this old post on the forum. It should provide you the answer:

http://www.cheresour...p-suction-side/

The important point is that with a negative suction pressure air ingress through a flange joint or valve packing will cause cavitation of the pump.

Regards,
Ankur.

[GS81Process]

Yes you can design a pump with a negative gauge suction pressure as long as the NPSHa>NPSHr. The negative gauge pressure number only signifies that the pressure at the pump suction is less than atmospheric pressure.

As Ankur mentioned, air ingress can be a problem since the pump suction operates below atmospheric pressure. However, air will not cause cavitation because air will not vapourize and then collapse back into liquid (assuming that operation is above air's critical temperature). Any entrained air in the system will reduce the capacity of the pump by occupying volume that would otherwise be filled by liquid. If the amount of air in the suction becomes excessive and is not bled out, then it can cause the pump to "air lock"- a condition where the pump capacity is reduced to zero.

I should add that I am also concerned by your statement that the pump suction pressure losses are quite large. Is your pump suction piping too small or is the route too complex with many fittings? In general, simple pump suction designs are better so as to avoid pressure losses and ensure adequate NPSHa.

Edited by GS81Process, 19 October 2011 - 03:19 PM.

[kkala]

We have often issued centrifugal pump data sheets with suction pressure of -0.05 Barg or -0.07 Barg, without any apparent problem. It is noted that in the data sheets LLLL of atmospheric tanks is considered, a rather unusual case for the tanks.

[pavanayi]

SAprocessGD,

Similar to the above posts, I have operated pumps having suction from surface condensers of steam turbines which are at 0.07-0.08 bar absolute, and discharing at 4 bara without any problems. As mentioned, block valves in the suction line had water seals in the glands to prevent air ingress.

[SAprocessGD]

Thanks all for your comments,

GS81Process: When I am designing pumps, I take into account worst scenerios. In my system, from our plot plan, location of our pump building to the furtest tank in the tank farm is quite long resulting in a pressure loss of about 10 kPag through the pipe. I agree that the pump system should be simple, and it is quite possible the layout of our system may be revised.

From what I am reading, I can see that a pump can have a negative suction pressure, so the pump suction is operating at pressures less than atmospheric. I understand a centrifugal pump doesn't "suck", so how does a liquid from a tank flow to a pump which has a negative suction pressure?

I'm just trying to understand the problem and I appriciate all the help in the matter. Cheers!

[GS81Process]

How does a liquid from a tank flow to a pump which has a negative gauge suction pressure?
It flows by pressure differential just like any other case. Forward flow
is possible until 0 kPaa not 0 kPag! Otherwise how would systems under vacuum (negative gauge pressure) be able to operate, such as that of a turbine exhaust as described by pavanayi?

10 kPa pressure drop is not necessarily excessive for a pump suction line. You say "pressure loss of about 10 kPag through the pipe" this statement does not make sense. Pressure drop is not a gauge pressure.

[kkala]

From what I am reading, I can see that a pump can have a negative suction pressure, so the pump suction is operating at pressures less than atmospheric. I understand a centrifugal pump doesn't "suck", so how does a liquid from a tank flow to a pump which has a negative suction pressure?
I'm just trying to understand the problem and I appreciate all the help in the matter.

Let us consider the suction pipe in this case, vertical for simplicity. Beginning of suction under Po = atmospheric pressure. Pressure of moving liquid at end of pipe = P<Po. Other forces exerted on suction fluid are its weight h*ρ*g*s downwards and friction F upwards (h=vertical suction length, ρ=density, g=gravity acceleration).
Equilibrium of forces on the moving fluid gives Po*s+h*ρ*g*s=F+P*s, that is F/s=Po-P+h*ρ*g (s=pipe section). As long as F>O, there is friction, thus motion.
On the other hand P=Po+h*ρ*g-F/s<Po, that is F/s>h*r*g. So F/s-h*ρ*g=Po-P>0. This shows that frictional force is more than the fluid weight, but it does not hamper motion.
Force (Po-P)*s+h*ρ*s pushes the fluid downwards, friction F is opposing. And moving fluid goes from Po to P, a lower pressure as we know. P*s is not the only force acting on the suction fluid.

Edited by kkala, 20 October 2011 - 03:59 PM.

[SAprocessGD]

Great guys! This is was really helpful.

Sorry GS81Process, I didn't mean to add the "g" on kPa.

Thanks again!

[ankur2061]

SAprocessGD,

For clarity on pump cavitation and NPSH refer the link below:

http://www.cheresour...ation-and-npsh/

Regards,
Ankur.