Determining pump flow from pump curve and pressure gauge

[mae1133]

We have an existing pump and pump curve. We know the physicals on the suction side of the pump, but the discharge is buried with no records of the physical layout, sizes, etc. We do have a discharge gauge installed about a foot above the pump discharge reading 16 psi. Knowing this, together with the pump curve, can the flow be estimated? Don't we also need a suction gauge to determine the differential?

 

[RVAmeche]

You need the suction gauge to determine based on dP of the pump. If you have a flow meter or control valve and are operating at a known flow, you can do the suction loss calculations, add the pump head, and check to see if the results match your discharge pressure gauge.

 

[mae1133]

That's what I thought. The operator has indicated we should be able to calculate flow with just the curve and the reading from the discharge pressure gauge, but I wanted to make sure I wasn't missing something before asking them to install a suction gauge.

 

[Artisi]

Alternatively, can you hire a doppler flow metre and check flow rate, inlet and outlet pressure are then not required.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[mae1133]

Certainly an option worth looking into, however, I'm not sure we have sufficient exposed straight lengths do install one.

 

[Artisi]

What's at the discharge end, can you fit a orifice plate for flow measurement

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[mae1133]

It would be quite difficult to fit an OP in the discharge piping. We have a reducer, check valve and isolation valve close coupled before heading underground. We are pretty limited to doing something on the suction side where we have about 4 ft of exposed piping. A suction gauge could easily be installed.

 

[Artisi]

Go back a step, why do you want to known the flow rate, and where will that lead if you know?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[mae1133]

Because we are adding another pumped system and tying into the existing discharge piping downstream of the existing pump. I need to determine the HGL of the tie in point to be able to select the new pump so I'm using the existing system to try to determine it.

 

[RVAmeche]

Is this to add flexibility (using existing pump or new pump) or to increase flow (using both together)?

 

[mae1133]

No. We are adding a new treatment process, but utilizing some of their existing infrastructure. Both pumps will not run at the same time. The existing system will become a backup if ever needed.

 

[RVAmeche]

If that's the case, then just make sure your new pump has at least the same pressure at the existing pump discharge tie point; assuming it's the same flow rate.

If you're increasing flow you need to do a calculation on the whole system from tank/supply to final discharge point and ignore existing data.

 

[LittleInch]

Even knowing the inlet pressure finding flow from the pump curve is not accurate (+/- 25%) at best.

There are plenty of insertion flow meters you could use in place of your pressure guage. Search "flanged insertion flow meter"

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[mae1133]

RVA: I do not know the existing flowrate. I do not know the details of the existing discharge as its buried.

 

[mae1133]

LittleInch: I can work with +/- 25%. I'm looking for an inexpensive and non-intrusive method. A strap-on FM may be my best bet.

 

[RVAmeche]

I would be very careful with qualifying statements when you propose the new pump will meet whatever criteria it needs to meet.

 

[mae1133]

RVA: I'm not sure what you are trying to imply, but my questions are not to "guess" at a pump selection but get some quantifiable data using the existing system to properly select a pump. The new pumps will be on VFD's, so I do have some flexibility.

 

[Orsiz]

What does the suction side look like? Is the fluid pumped out of an atmospheric vessel? Do you know the static head? Is the amount of suction piping and corresponding fittings small?

Based on this information, the suction head can be estimated. As you know the discharge pressure, the TDH and hence the flow rate can be estimated from the pump curve.

 

[mae1133]

Orsiz: Yes, tank is open to atmosphere and roughly 6 ft of liquid (5 ft above pump centerline). The piping is 3" SS and roughly 4 ft long with one 90 elbow and one ball valve. So, yes to your questions. You are right, suction should be slightly less than 5 ft or 2 psi+/-.

 

[Artisi]

OK, in my experience field testing even under ideal conditions is very arbitrary and inconclusive. At the moment you know the discharge pressure only, even knowing the inlet pressure could give you some idea of the total head, but to what degree of accuracy.
You can measure the input power to the motor, and with a lot of if buts and maybe's assume the motor efficiency, from here you can guess the pump input power and look at your pump curve - is this a test curve or a generic published curve for this particular model pump - now, is the pump in the same mechanical state on which it was tested (assuming it was) running at the same speed etc.

Lots of unknowns and guesses as a field test can never reproduce a test result and if a generic performance curve the field test could be near useless - of course this depends on what accuracy you want.

I have experienced a number of pumps rejected by site tests as not performing as quoted / sold etc. but on return to the manufacturer and retested under test conditions they perform as stipulated.

The only way to establish accurate flow is to measure it.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)