Non Repeatability of Turbine Meters

[Rabisaab]

We have a Smith Guardsmann "G" Series Turbine meter. It transmitts the pulses to a High Speed Counter of Allen Bradley PLC. The meter is brand new. We need to prove it However during the proving process, we are not getting it repeatable. (100 litres difference per 5000 litres set volume)

Kindly if somebody can advice how to overcome this proble.

 

[BigInch]

Is that a constant error you get every time you run 5000 l through the prover, or is the measurement varying within +/- 50 liters each time you run 5000 l through the prover?

http://virtualpipeline.spaces.msn.com

 

[Rabisaab]

The measurement turbine meter is varying within +/- 50 litres.

 

[BigInch]

OK, lets see what else is happening.

So you are trying to do the initial onsite proving and you are getting variations in the K factor (pulses/liter).

What's the K factor listed on the meter factory test card and what are you getting onsite?

Are you only using one product during proving?
Will you run different products later?

Were there variations of pressure or temperature that happened during a prover run? ie, was pressure and temperature constant from start, during and to the end of each prove?

Were there variations of pressure or temperature between the runs showing the variations, or all runs were done at the same P & T. If P or T changed, by how much?

Was flowrate varying during or between proving runs?

Are you near the products vapor pressures?

http://virtualpipeline.spaces.msn.com

 

[Rabisaab]

I am proving each meter with the operating product. I have set the K-factor as the factory K-Factor but I am not getting it repeatable.

The temperature and pressure are the same during the proving process, however the flow rate is not constant as when we start the batch the Flow Control valve will open gradually and the same when we are finishing the batch the flow control valve will close gradually, but this must not have any effect on the repeatability of the flow meter.

I am doubting the counting of pulses at the PLC end. The PLC is counting the frequency of pulses (rate) and not counting the pulses itself. May this have an effect on the repeatability of the turbine meter readings?

 

[BigInch]

Right, the control sounds adequate.

If you don't get a count of the total pulses from a totalizer function on the PLC, how do you know you ran 5000 liters? Are you counting pulses using some other device? Is there a totalizer function on the PLC? Was the PLC also from Smith? What I'm asking is who programmed the PLC?

It kinda' looks like it is the PLC, but if the problem is is in the meter itself, you're going to have to disassemble and inspect it. Dirty bearings, a loose stator, or (do you suppose) a missing blade could all affect the total number of pulses.

http://virtualpipeline.spaces.msn.com

 

[Rabisaab]

Yes the PLC has a totalizer. The supplier of the PLC is the manufacturer who supplied the control panel of our control system. Also the Smith turbine has a local totalizer which is independant from the PLC totalizer. Now the PLC totalizer is not repeating. The PLC is using counter rate and not pulses counter. Do you think this is causing the problem?

 

[BigInch]

Perfect example of what I always say, "A man with two watches never knows the time."

So then how's the local totalizer doing. Is that one running OK?

I'm getting out of my depth with programming PLCs, so I can't say for sure, but I should think it would work better with a count of total pulses since last reset. If it counts pulse rate, there's always a possibility that its dropping volume bits when it integrates over time steps to reach a total.

http://virtualpipeline.spaces.msn.com

 

[pipehead]

PLC's ARE NOT totalizers, they are averagers.

 

[Ashereng]

I would suggest trying to count the pulse rather than the pulse frequency. Can't hurt.


pipehead,

Can you expand on your comment? What do you mean that "PLC's are not toalizers, they are averagers."? A total is a summation. A PLC can sum quite well in my experience.

"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[CostasV]

Could you give us your volume data
theoritical volume - measured volume
for all of your test?

Costas

 

[jmw]

How about the fluid viscosity?
Smith do supply viscometers and offer a viscosity correction if required.
Do you have any entrained air?
Is the installation according to the manual i.e. do you have the necessary straight pipe lengths before and after the meter and does your pipe design avoid generating swirl at the inlet to the meter run?
Are you within the linear range of the meter? I assume since you quote a k-factor rather than a factor set that you do not have a linearising fit, or am I wrong?
Do you have a flow rate display and if so, how stable are the readings?


JMW

http://www.ViscoAnalyser.com

 

[JLSeagull]

Entrained air or vapor would cause lots of inconsistency.