flow meter caliberation

We calibrated propeller flow meters in the 70's and devised this volumetric method.

We had just acquired a magnetic flowmeter (with the claim of very accurate) but did not use it because of magnetic interference from the parsing railway line.

equipment

A constant header tank - pumping in the tank and have it overflow the surplus.

Masonry tank (I guess 20 x 3 x 1.5 deep) Our volume were rather accurate and we did check for irregularities.

Stop watch and some mechanical water level meters with vernier scales.

method

Put about 5 meters in a row with ample spacing between them, and we did rotate the meters.

Fill the tank to the zero value.

Take all the meter readings.

fill the tank to the predetermined level (if I remember correct just under 50m3)

Take readings again and calculated the accuracy.

Result :- most of the tested meters did not reached their claimed accuracy with clean water. (up to 50% errors were recorded.)

There is ISO standard for calibration of flow meters ISO 4185. There are other methods also for calibration of flowmeters. These methods give calibration to 0.1% of actual flow. The rigs designed are accrdited by national accreditation bodies.

Hendrik says, most of the tested meters did not reached their claimed accuracy with clean water. (up to 50% errors were recorded.)

This is like checking the accuracy of Micrometer with a foot scale.

The method Hendrick used has no validation, calibration for each instrument used. The volume between Zero Level and predetermined mark needs accurate calibration, the stop watch needs acurate calibration, many other factors need to be taken in to consideration line, temperature of the water, volume of measuring tank at that temperature. It is quite invalved thing to calibrate the meter.

It is just impossible to calibrate the meter at site, just before installing.

Better you visit and witness the calibration at manufacturer's facility.

Dear Friend,

I suppose that you have a very tough client/ consultant!

Anyway, the query which you have is all about instrumentation (Magnetic Flowmeters).

From the manufacturer it is already pre-calibrated, but client sometimes requires Bench Calibration before installing the unit.

Then, when installed on the pipe, you need to do the on-field calibration and pre-commissioning calibration.

For all this series of calibration, all you need is a VOLT/mA Loop Calibrator but certified by any third party calibrator. Just feed 4-20 mA (4mA_assuming 0 or mininum flow and 20mA_for maximum flow).

_commissioning engineer_

Dear Guest, It is a good idea to install a "pre-calibrated" instrument. However, once installed it will need final "calibration". You seem to be asking about the pre-installation calibration method so the information I give here may not be of that much interest to you but it may be useful to you at a later stage.

When people talk about "calibration" they typically are referring to the process of comparing the actual reading (say flow in your case) to a known reading, then adjusting the measuring instrument to give the same value as the know reading (+/-) and then comparing the actual reading to a known reading again. This would be done over a range of values.

In the case of the flow meter, you should install it in such a way that a know calibrated flowmeter can be put into the same line and then you can reference the actual reading from the permanent flow meter and the reading taken from the known calibrated flow meter. This method is not possibly in all situations, take for example a water loop, in alot of cases it is not possible/ desirable to break open the loop and install a second flowmeter. If this is the case then the method to calibrated the instrument needs to be designed into the system.

Kind Regards

Mr. W.A Snow

it in such a way that a know calibrated flowmeter can be put into the same line and then you can reference the actual reading from the permanent flow meter and the reading taken from the known calibrated flow meter.

Then question comes, how to calibrate this KNOWN Calibrated meter? Thus again it is precalibrated. If you believe this precalibrated meter, why not believe in your original meter?

You can field calibrate the meter, using temporarily fitted Prover in series. But this method is too costly.

All other methods of knowing the KNOWN reading to compare with actual reading are inaccurate than the actual reading. (Temonology used is as you have used in your para 2)

Dear Gsuhas, Perhaps you, I or both of us have misunderstood what was being requested. My understanding of the original post was that this person wanted to install a flowmeter (in this case) and they were asking about the calibratin of the unit. What I was responding to was the ongoing calibration of the flowmeter, say one year after installation. The "Known" calibrated flowmeter would be calibrated by an acredited body and would be of a higher accuracy than the unit that this poster was trying to calibrate. My point was that the design of the system should allow for the calibration of the instruments post installation - how many systems have you seen whereby the instruments are totally inaccessible and there is no method for their calibration ?

I hope this clears up my though process and if I am incorrect then please accept my apology.

Regards

Mr. W.A Snow

Hello HIS/WAS

You are right. Your point is clearer now than your original post.

I would like to add some thing more. Now a days, all the manufacturers have calibration laboratory, which is accredited. In this situation, when manufacturer lab itself is accredited body and the original meter supplied (which we want to re-calibrate after say one year) is calibrated in accredited laboratory. Now you propose to use one more flow meter calibrated in accredited laboratory (thus again in same laboratory). Being calibrated in same laboratory, the reference BMC (Best Measurement Capability) of the laboratory remains same. This will mean in turn, we will be calibrating old flow meter with reference to recently calibrated flowmeter. Yes, this means something.

Now if recently calibrated flowmeter is better in accuracy / uncertainty by 1 order then it is very nice and really meaningful. If not it has no great advantage.

But if the new flowmeter is calibrated in some accredited laboratory with much better BMC, it is really means something. Calibration in just accredited laboratory has no great meaning. (this is what I meant by using prover)

I hope my point is also clearer and certainly you will agree with me.

We both mean same thing, but narration is probably different.

Regards

it is not costly to to calibrate in the field. By contract we must calibrate our meters monthly. The cost is quite low. At the meter we install a block valve with 2 by pass vales, 3 new valves. The portable is is brought in and for $500 to $1000 US, they prove the meter. It takes about 4 hours. Production is not interupted.

Dear Vicini, the question is not about the cost or the interuption to production. The question is how i.e. what class of instrument do you use to calibrate the instrument in the field. When you are calibrating a sensor in the field, where and to what standard is the "portable" instrument calibrated (hopefully you will see what I am getting at). In my experience the "portable" instrument would be calibrated by an accrediated laboratory, probably because it is more cost effective to do it this way. I would be interested in hearing your thoughts/ experiences on this.

Regards

Mr. W.A Snow

To calibrate a portable ballistic or ball prover, the mechanical portion is proven agaiaist an NIST traceable volume container. The unit is filled with water and compared to the standard. This is done annually

The system has adjustments for temperature and pressure. Those devices are sent to a NIST traceable lab for calibration. You calibrate with a device that has at least 2 times the resolution. These are recertified annually too.

You can calibrate the meter with another meter that is exactly the same make and model too. The reference meter is calibrated in the lab or agaianst any other meter system listed above. This meter can be taken to the site and installed in series with the meter you need to prove.

For highly sensitive metering, I installed two meters in series and then monitored the difference. If the meters diverge outside SPC limits, then one is ent back to the lab for calibration. Or if there is sufficent NIST traceable calibration devices, it can be calibrated on site.

Dear Vicini, thanks for your response and the detail. This is how we currently perform our calibrations i.e. we do not certify the calibration instruments ourselves we send them out to an accredited laboratory (NIST traceable). On some instruments we send them out twice or three times a year - I am not really sure if this is technically required or if there is some other reason for this. I must actually try and find out. Would anyone like to hazard a guess as to why this would be done (Its not a homework question!!!) ?

Regards

Mr. W.A Snow

the frequency or recertifying is driven by three things. 1 Contracts, we have contracts that are specific. 2 historical data that has been statistically anaylized. 3 Unwritten Contracts, yes again we do things because that is what our mothers did, and their mothers before them.

Dear Vicini, I checked on this and

Number 1 Contracts: Not directly applicable to us.

Number 2 Historical Data: The data would support a less frequent calibration interval.

Number 3 Unwritten Contracts: This is in fact the reason we do it. "We have always done it this way and it has served us well" / "Its not broken so don't fix it". I think the intention was to start off with a short calibration interval and then use the data to support an extension (good logical approach), however somewhere down the line it became "the norm" to have this calibration interval.

Regards

Mr. W.A Snow

great engineering investigation. Dr Demming would be pround.

under API standards you can put a 3 valve maifold and calibrate you meter insitu with ANY bench calibrated meter in series. BUT, bring in a potable prover with an NIST/API traceble meter to put in series with your meter on a regular basis. These portablr meters usr a Ballistic or ball meter meter.

Any calibration must be carried out in a controlled and well defined environment. Any other calibration (as you call it) can only at best called a functional test.

Take the flowmeter to a standard calibration laboratory with a prover and get it calibrated and that is it.

Any other method used will only lead to discrepancies at a later date. More so if the flow meter is used for commercial purposes.

You live in a dream world of the research and labs. I calibrate meters in the field that are 100% NIST traceable. These meters transfer $2,000,000 per day of products to customers. They are repeatable to .50% and accurate to .50%.

We send some of our meters to proving stations out in the middle of a field in Iowa with a small tin building. We do this every 5 to 10 years.

The largest single meter I ever calibrated moved $250,000 per hour! No lab on that one, all in the field.