ASME B313 Hydrotest gauge tollerance 3

[graycwi]

When performing hydro static testing of process piping B31.3, is there a tolerance for gauges. For example: The manufacturer is using two gauges; one reads 1000 psi and the other gauge is reading 1020 psi. The gauges are being re calibrated every 6 months however some of the gauges do not read the same while under pressure for the same test. Thank you in advance.

 

[GenB]

use digitals.

 

[TGS4]

There is no requirement to use multiple pressure gauges during the hydrostatic leak test. Therefore, I would recommend not using more than one.

This reminds me of the saying: the man with one watch always knows what time it is, while the man with two watches never knows what time it is.

 

[moltenmetal]

You are attempting to read the "gauge" beyond its accuracy. 1000 and 1020 psig are the same if the accuracy is +/- 1%. Back in the old days you would have mirrored scale "test gauges" with accuracies better than +/- 1%, at least in theory...As to a digital "gauge", it of course gives you "digital display bias", tricking your brain into thinking that all the figures displayed are significant when they absolutely may not be!

B31.3 has no requirements for the accuracy of test gauges. Your AI, if you have one, may point you to some and expect you to follow them. Here we have a procedure for checking shop gauges against a master transmitter which is sent out for calibration yearly. That satisfies our AI. The gauges we use are probably closer to +/- 2% within the measured range.

 

[weldstan]

Most analog test gages have an accuracy of plus or minus 1% as defined by the manufacturer. Take look at the manufacturer's accuracy statement for the gages.

 

[moltenmetal]

+/- 1% of the full scale reading, whereas your reading typically is between 1/3 and 2/3 of the full scale. Do the math- the gauge might be +/- 3% of your reading.

 

[jte]

Good point, moletnmetal.

Reminds me of a few years back when I had to change our corporate spec's as a result of a good discussion with a high quality fabricator. I also get frustrated with "digital bias" in issues like reporting a vessel's fabricated weight using 7 significant figures for a roughly 200,000 pound (or kg, your choice) vessel. So the weight changes when a seagull lands on it. Just because we have a readout with a lot of significant figures does not mean the result is precise.

 

[Amir35]

There is no requirements in B31.3 Code book but respected companies have their Specification as following:
"Pressure gauges shall be of the Bourdon tube type, complying with BS 1780. The nominal diameter of each gauge shall be 100mm.dial gauges used for testing shall have a range of a about double, but not less than 1.5 times the test pressure, in no case shall be less than 1.5 nor more than 4 times of intended maximum test pressure."

 

[bpv66]

Using one gauge is great until you send it to the test lab for annual calibration and find out it was reading 20 psi too high. That calls into question all hydrostatic tests performed using that gauge since last calibration. I wouldn't want to disposition that NCR.

I always recommend two gauges when testing to ASME standards.