API 653 joint efficiency

If the tank standard API 650, yes .. BUTT joints..



Use it up, wear it out;
Make it do, or do without.

NEW ENGLAND MAXIM

Hi HTURKAK, thank you for you answer
but in the paragraph it is confusing.
It says: welds or joints
It does not specify butt welds only.
I want to know if I can use E=1 for corrosion detected near the bottom-shell joint weld.
Corrosion is located less than 25 mm (1") from the bottom-shell weld.


Now I have another question.

The paragraph says: Use table 4.2 if the original E is unknown.

Suppose a tank was built according to API 650 in the year 2000. I do not know E. What value of E should I use to calculate tmin for a complete course if I have no further information on Applicability or Limits. the most unfavorable one?

Best regards

-The definition of E ( efficiency )
E is the original joint efficiency for the tank. Use Table 4.2 if original E is unknown. E = 1.0 when evaluating the
retirement thickness in a corroded plate, when away from welds or joints by at least the greater of 1 in. or
twice the plate thickness,

-Table 4.2 (Joint Efficiencies for Welded Joints) is applicable for API 650 tanks with butt joints, API 12C tanks with butt and lap joints and for the tanks with Unknown standard butt and lap joints,


- If you don't have as -built documents ( test documents e.g. RT tests performed as per Annex A ) the applicable Joint
Efficiency E =0.7

I will suggest you to provide more info. for the subject tank ( sie , content , service temperature, shell thicknesses and corrosion survey ) to get better responds . The pictures implies the replacement of Tank Bottom Plates ( as per 9.10.2 should be an option.




Use it up, wear it out;
Make it do, or do without.

NEW ENGLAND MAXIM

Let's go back to basics. In this case you are evaluating the retirement thickness of the steel. You are computing the required thickness of the steel using the product load which is resisted by hoop tension in the shell plate (think steel bands around old wooden barrels). The tension in the plate is resisted by the full thickness of the material. Joints are the weakest part of this or almost any structure. The joint can be butt welded, lap welded or riveted. Each of those types of joints is characterized by a joint efficiency (E). If E is 100% then there is no loss of strength at the joint. If E is less than 100% then the steel must be increased in thickness in this area to carry the same load as the rest of the plate (E is in the denominator of the formula). Since tanks are made with single thickness plate (unlike bicycle frame tubing for example) the entire plate will be thicker because at the joint the tank is weaker. With a joint efficiency of less than 100% (or 1.0) there will more steel away from the joint than you need to resist the tension from product loading.

With this background you can see that the joint efficiency is the evaluation of a shell vertical joint which can be in several configutations including welded. Other welds do not matter unless they also reduce the ability of the shell steel to resist tension. Fillet welds around a nozzle are not considered to reduce the ability of the shell plate to resist hoop tension. Once you are an inch away from corner welds, the corner weld is no longer considered to affect the ability of the shell plate to resist tension.

If the original joint efficiency is not known from original documents or testing that you do, API 653 says to use table 4.2 when calculating the thickness of the complete course (or ring) of the shell. But as you know and API 653 reminds us, joint efficiency only matters when you are evaluating the shell thickness at a joint. If you are not at a joint or near a corner weld you don't have to increase the calculated required thickness by considering a joint efficiency. Away from a joint or the corner weld you can use E=100% or 1.0.

IFRs said:

Joints are the weakest part of this or almost any structure

Click to expand...

It isn't true.
RT is an effective inspection tool and is considered an aid to quality control.
The E does not imply that the weld is of greater or lesser resistance, it only demonstrates the degree of inspection that is carried out.
Let's assume a sheet of 10x2400x15000 mm, which I then roll and get an average diameter of 4775, and a butt weld.
Depending on the E that defines this cylinder, it can be used for 3 different design pressures. But the quality of the welding is always the same: it has no defects, obviously.
There are workshops that always weld very well (automatic). The quality is the same with E=1 and with E=0.7.
The difference is in the inspection level with RT, but that same weld with E=0.7 resists the same pressure as with E=1
​
Regards

Since CJP welded joints are expected to have 100% efficiency my statement is not true and I thank you for correcting me.

If a welded tank is uniformly loaded to failure, I would expect it to have a greater probability of failure at or near a discontinuity, change in geometry, lamination, incipent crack, altered chemisty, change in hardness or ductility, etc - I think all are more likely to occur at the joints.

No material is perfectly homogenous and uniform in properties but at the joints I think it is less so. In this respect the joints are the weak points and deserve additional NDE.

I also find that new engineers tend to overlook the joints, may not understand how they work and/or leave the details to someone else.

I would however dispute the contention that all butt welds are of equal quality and strength and the NDE done only is something on paper. I think the degree of NDE in some sense does imply and sometimes reflect the actual quality of the weld and certainly is used for a level of confidence in the joint.