orifice plate versus venturi tube

[CH5OH]

in a pump installation, where the flow is measured in discharge line, using an orifice plate, would it be benificial to replace orifice plate by venturi tube?
less system back pressure=less pump energy required...
Could the losses one versus the other be expressed in percentage?

 

[MartinSr00]

The Venturi would have lower losses, because the energy loss is small. Typical discharge coefficients (C) in the .984 range. In essence, this means 97% (.984 squared)of the energy is recovered between the inlet and the throat of the venturi.

So, there's surely a little more loss in the downstream section of the venturi. Maybe it's just as much as on the upstream. So, let's say another 2%. Loss is say 5% or so tops if your venturi surfaces are not craggy.

To determine the losses in an orifice plate, check the ASME Standard MFC-3M. The C value for an orifice plate is 0.5959+0.0312xB^2.1-0.1840xB^8+91.71xB^2.5xRD^(-.75) where B is the diameter ratio d/D, and RD is the reynold's number of the pipe. This is, I'm nearly certain, empirical.

Since C is proportional to the flow, and the energy loss is proportional to the square of the flow, the ratio of the C's squared should be the percentage you want.

But in rough terms, let's say the venturi is almost lossless - that is, the pressure difference between the inlet and the troat of the ventri is recovered in the downstream section. If you measured the pressure difference from inlet to outlet it would be quite small. The orifice, the water horsepower loss would be rho x g x deltaH where rho is the density of the fluid, g is the gravitational constant, and H is the head in feet. In English units, that's Q*deltaH/(3960*sp. grav), where Q is in gpm and delta H is in feet.

This is probably more than you wished to know. Please check my math.

 

[CH5OH]

thanks,
would it be correct to state:
orifice loss (watt)=deltaP (pressure over orifice in Pascal) multiplied by flow (m3/s)
?

 

[MartinSr00]

Sorry for the late reply.

The pressure loss stands in for head which is work/unit mass, the flow is mass/unit time. Multiply them that way, and it's work/unit time - power.

You can do it with deltaP * flow * density * a "fudge" factor. That's what I did above with the 3960 constant and english units.