Magnetic Flow Meter anomalous reading

[dbill74]

Got a magnetic flow meter installed on a reverse osmosis (RO) system and getting the occasional spike/drop in flow reading. That is to say when looking at logs, we are seeing a reading where the flow rate doubles, then drops to about half the normal flow rate. There is no ramp up/down to these flow rates, just a brief reading of excessive/low flow, before the readings return to normal.

The flow meter is downstream of a centrifugal pump with an output pressure of ~800 pisg. This water has NOT passed through the RO membranes yet.

What can cause a magnetic flow meter to spike like this? We are trouble shooting other issues and it has been suggested that air is causing our problems, could an air pocket cause this apparent spike in flow?

Pump is controlled via VFD.

 

[Compositepro]

Likely an air pocket.

 

[ashtree]

Magnetic flow meters don't read air. If air was somehow involved it would have to somehow hold back the flow which releases suddenly and causes the spike and you will then have a period of low flow behind it. Is this what is happening?

What happens to the all pressures on the RO during these spikes? If the pressure spikes as well it certainly confirms that it is an actual spike not just some instrumentation issue.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[gerhardl]

A common fault for any flowmeter is placement too near a source for non-linear flow or disturbance; bend, pump etc. Is the installation and flow characteristic as recommended by the manufacturer?

 

[bimr]

If this is a new installation, make sure that the flow meter is electrically grounded and bonded. I witnessed what you have described on a magmeter installation.

installation-grounding-of-magmeters

Once a system similar to what you have described has been properly commissioned, it is not likely that entrained air will be present.

Magmeters are generally not susceptible to improper placement near a source for non-linear flow or disturbance; bend, pump etc.

"Magmeters are relatively insensitive to errors caused by nonsymmetrical velocity patterns or swirl. The general rule of thumb for straight piping is a five-diameter length of piping upstream and three diameters downstream from the meter (measured from the center of the tube). Independent testing has shown that magmeters may be affected by piping effects when the length of upstream straight pipe is less than three pipe diameters. Errors from piping effects generally run between 0.1 percent and 1.5 percent, depending on the exact configuration of piping and length of pipe run upstream of the meter."

Link

 

[dbill74]

Thank you for the responses.
To answer a couple questions:
[ol 1]
[li]The system and meter are a little over a year old.[/li]
[li]It is measuring flow direct from a 3-stage high pressure pump (RO booster pump). From the pump there is ~4 feet of 4" steel pipe, transitions to 6" dia. for ~2 feet, then there is the meter. 1 foot behind the meter is a Tee in which another flow (from another pump) combines with this pump's flow before going to a bank of RO membranes.[/li]
[li]Although I have yet to see logs (they have been requested), it is my understanding that this anomaly is occuring once every 4-6 hours.[/li]
[/ol]

 

[John2025]

WAG, some kind of disturbance from the other pump?

 

[bimr]

Is the magmeter grounded? Grounding and bonding the magmeter is very inexpensive and will most likely fix the problem.

"Proper grounding of a magmeter is one of the most important installation details."

installation-grounding-of-magmeters

"About 80 percent of the time, problems with the flow measurement readings provided by a magmeter can be traced directly to improper solution ground connections."

www.flowcontrolnetwork.com/magmeter-grounding-strategy

 

[djs]

I have also encountered a similiar problem with a loose connection between the electrodes and the transmitter. Check all connections for tightness.

 

[ashtree]

I agree with bimr but it may also be an electrical of RFI induced spike on the cabling. Check what else happens in the plant at the same time. It could be a motor stopping or starting, a valve operating etc that is inducing the spike.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[dbill74]

Thanks for the input everyone. Will definitely have to check the grounding and wiring connections.