Pumps and entrance, exit effects

[rocketscientist]

A question came up today regarding pressure drop calculations around a centrifugal pump. One gentlemen says that there is no need to add a K factor (K = 1, typically)for a pump discharge. He says that the pump supplier accounts for the pressure drops at the inlet and outlet of a pump.

What do you say? Are you supposed to add entrance and exit effects in a hydraulic calculation with a pump or does the pump curve already account for these losses?

 

[stanier]

I have never added a K value for such effects at the pump in 40 years and have yet to have it bite me where it hurts.

"Sharing knowledge is the way to immortality"
His Holiness the Dalai Lama.

http://waterhammer.hopout.com.au/

 

[rmw]

Your pump performance curves are stated from inlet to outlet - from centerline of inlet to centerline of outlet. All effects inbetween those points are accounted for by the pump vendor.

If not, find a reputable vendor.

rmw

 

[BigInch]

Meeting the NPSHr curve is sufficient as that provides a velocity head in the suction high enough to overcome minor losses and allow the pump to generate its predicted differential head as advertised.

We will design everything from now on using only S.I. units ... except for the pipe diameter. Unk. British engineer

 

[hydromech]

For water pump supplying cooling for HD diesel engines, we measure against the customers specification at the pump outlet. If the pressure at the outlet of the pump meets the customer requirements, that our job done.

The pump is designed to generate a pressure differential between the inlet and the outlet.

If the customers makes a bad job of the engine block and stuff up the water flow out of the pump, that is their problem. We do try to educate our customers, but they seldom listen.

Hydromech

 

[rocketscientist]

These are all good additions but I am trying to calculate the head required for the pump based on the pipe network. Part of that network is the exit and entrance of the pump.

In most cases, I add entrance and exit effects. Some engineers add them only if the flow goes from a large plenum to a small one. I typically add a K=1 for exit and K=0.5 or 0.25 for an entrance. I do this for heat exchangers, for example because the fluid is distributed to a tube bundle; there are a lot of un-accounted for pressure losses there.

From what I gather, the pressure drop through the pump is accounted for, which makes sense. This makes pump unique, in my opinion.

The other reason why I am generous with entrance and exit effects is because the simple Crane K factor method is un-conservative. I figure with equivalent lengths over-estimating pressure drop and Crane under-estimating it I can hedge my bet with entrance and exit effects.

Thanks for your input.

 

[BigInch]

The pump is a bit different than a totally parasitic load where everything is a loss, thus rendering the only important quantity the net value, hopefully a gain in head.

By the way, do you add the inlet velocity head at the pump and subtract it from the discharge, or do you ignore that?

We will design everything from now on using only S.I. units ... except for the pipe diameter. Unk. British engineer

 

[rocketscientist]

Good point; this would make sense if you want the head across the pump. Doing this would reduce the head across the pump, but only slightly.

 

[Pumpsonly]

The One gentleman is correct. The pressure drop at the inlect and out let of the pump is taken care of by the pump manufacturer.
You take care of the losses right before the suction flange and right after the discharge flange.This is true for all the pumps having a suction flange and a discharge flange.

For pumps without a suction flange,such as sump pump and vertical submerged pump, the definition become slightly more complicated.

The difference in velocity head , if any, between the suction and discharge is included in the computing the TDH during the pump test.

 

[katmar]

To answer your question you really have to focus on the question "what is an entrance or exit effect?"

An entrance effect is where the fluid is conveyed from a large diameter to a much smaller diameter. An example would be the outlet from a storage tank where the fluid has to be diverted into a much smaller diameter than when it was in the tank. So, for a pipeline like this that starts with an outlet from a tank you would include an entrance effect. In a pump that has a pipe connected to the suction flange there is no such change in diameter at the pump and no entrance effect needs to be taken into account. As others have already mentioned, any diameter changes within the pump will be included in the pump curve.

If the pump suction were bolted directly to a tank outlet there would be an entrance effect if you want to take the pump's suction pressure as the hydrostatic head in the tank.

An exit effect is a little more complex, because it actually does not occur at the exit. The exit effect takes into account the kinetic head of the fluid that was imparted to the fluid long before the exit, but which is lost (or, more accurately, not recovered) at the exit if the kinetic energy cannot be converted to pressure energy as per Bennoulli. Again, at a pump discharge there is usually a continuation into a pipeline (perhaps via a reducer) and the kinetic energy is not lost (some may be lost if there is a reducer - you would have to apply an applicable K-value for the reducer). The pump curve shows the total dynamic head, which includes this kinetic energy. Of course if the pump discharges directly to atmosphere all the kinetic energy is lost and there is no pressure component.

Katmar Software - Engineering & Risk Analysis Software

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[BigInch]

That would be a correction that technically is applied to a component, not part of the pump, ie. the tank's outlet.

We will design everything from now on using only S.I. units ... except for the pipe diameter. Unk. British engineer

 

[katmar]

BigInch, yes that is what I meant. If the pressure across the pump is taken as being between the hydrostatic pressure in the tank to the pump's discharge then the entrance effect in the tank's outlet becomes part of the pump system. Alternatively you could calculate the pressure drop through the tank outlet and correct the pump suction pressure for that loss.

Katmar Software - Engineering & Risk Analysis Software

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"