Design thickness + Corrosion allowance - Over grinding 1

[Syahar1975]

Good morning Gents,

This issue always happen in tank construction in my past and current projects.

The tank designer by calculation will have the shell design thickness. Then based on project specs the design thickness will be added by the corrosion allowance. Sometime the number is odd let say 7.89 mm. The order to steel mill would 8 mm.

During construction many time happen our grinder or helper than learning about grinder will over grinding when they are removing the temporary attachments.

Yes, sometimes we can ask the welder to build up that over grinding area. But also happen we miss to see the other over grinding area.

It can be seen clearly after the tank is painted.

The QC inspector then will point out that area are over grinding so the thickness of the plate is reduced by 1 or 2 mm.

So far I always win by giving the above reason design thickness + corrosion + round up.

How do you see this ?

Your comments is highly appreciated.

 

[e43u8]

Syahar1975,

Where you account on round up as a margin for thickness selection, 0.11 mm wouldn't be a reasonable value. For the shell courses with such low round up, it's better you would select higher standard thickness after round up to have adequate room for prbable thickness reduction in various manufacturing steps...

 

[fegenbush]

Per 4.2.1.2.3 of API 650, an under tolerance of 0.01 in (0.3 mm) is permissible on plate material. If your grinding reduces the thickness by more than this amount beyond the calculated thickness, rework is required to be in compliance with the standard.

 

[Syahar1975]

Thanks for the respond.

e43u8,

That is an example only. In reality the tank designer will give us the order thickness and mostly even number.

fegenbush,

Yes I knew API 650 give is 0.01 tolerance. This is also the argument given by our QC inspector.

But again I always "win" based on that design + corrosion allowance.

 

[JStephen]

I would take this as a situation that simply isn't covered by the standard.
There is an underthickness tolerance, but that is applied to the edge of the plate, and determines plate acceptability, not workmanship.
I don't think it's the intent to go check thickness at every grind mark to make sure they all meet that minimum.
I believe API-653 has criteria for localized thin spots due to corrosion, and that could be applied here, although not intended for that purpose.
Also note 7.1.6, where actual tears/holes are repaired by welding and grinding.

 

[Duwe6]

As jstephen alluded to, there is a methodology to compute locally thinned areas in API 653. Use it.
Weld overlays tend to cause 'puckering' of the plate; adding insert plates usually makes the insert go flat [vs. following the tank's curve] You have to ask "Is the repair going to be an overall improvement of the defect, or will it exchange one defect for a different one?"