Sample size ultrasonic testing

[Klikoos]

Hello everyone,

The company I work for has a containment vessel with an area of roughly 25.000 ft^2 (or 2300 m^2). Right now they do 4 spots per plate which seems a bit excessive to me.

Would the formula; "margin of error = standard deviation divided by the root of the number of measurements" be applicable when determining a safe amount of measurements?

Right now it doesn't hold anything and there is little to no local corrosion, also the wall have a uniform thickness.

 

[compit]

If you look to API 653 you will see that the quantity is laid out per section through a tank, not per plate and is determined by the diameter of the tank.

I don't really understand where your margin of error figure comes from nor you formula but I can say that the time to take a single measurement on a clean plate is very quick, a couple of seconds, so tanking 1 or 4 should not have any real impact.

What is more critical is where on each plate each measurement is taken, and the ability of the ultrasonic inspector to search for pitting in otherwise non-corroded plate. - Something I could o into in detail.

I hope that helps somewhat.