API 650 Failure Pressure uplift

[mariog123]

I kindly ask your interpretation of the "Uplift Load Case/ Failure Pressure" in Table 5.21—Uplift Loads.

By note a. in table Table 5.21, this case is valid for special cases: "Failure pressure applies to tanks falling under F.1.3 only. The failure pressure shall be calculated using nominal thicknesses."

F.1.3 stipulates "F.1.3 Internal pressures that exceed the nominal weight of the shell, roof, and framing but do not exceed 18 kPa (21/2 lbf/in.2) gauge when the shell is anchored to a counterbalancing weight, such as a concrete ringwall, are covered in F.2 and F.7."

In my understanding, F.1.3 refers to tanks having uplift in operational and/or design pressure. By API approach of frangibility, a frangible tank has the have no uplift in case of roof failure. It appears that a tank which has uplift in operational/design pressure (before roof failure) cannot be frangible, because the tank shall be certainly subject to uplift at roof failure pressure. I would conclude that F.1.3 refers to non-frangible tanks, however the current interpretation is that the uplift case with 1.5×Pf refers to frangible tanks, isn't it?

And second, which is the meaning of the uplift case with 1.5 × Pf? I expect to have roof failure at pf, but still I can apply 1.5×Pf as internal pressure?

Thank you and best regards.

 

[JStephen]

There are several different conditions given under which a roof can be considered frangible. One of those is where the roof is weak enough, and the shell heavy enough, that the roof will fail prior to pulling the shell from the bottom. Another condition is if the tank is adequately anchored so that the roof will fail prior to lifting the shell/foundation.

However, even in cases where the roof is not considered frangible, API is still requiring the anchorage to be stronger than the roof failure mechanism. I don't know the entire thinking here. One issue might be that in the case of combustion in the tank, you'd want the roof to fail rather than the corner weld, but API seems reluctant to indicate that as a design issue. The 1.5Pf item might be in there for that reason.

 

[mariog123]

JStephen,

Thank you for your opinion.

I would conclude that "1.5*pf" addresses to non-frangible tanks, despite the common opinion which is opposite.
And I would understand that a non-frangible tank (subject to F.1.3) has a good chance to be frangible in behavior, even remains non-frangible as API classification.

By the other hand, it appears that API 650 considers effectively "1.5*pf" as a load case when the tank is under F.1.3.
Not only "1.5*pf" is considered as "limit case" for anchorage & shell at anchor attachments, but also F.7.4 c) should be considered for the counterbalancing weight.

Best regards.