API 650 Appendix A weld spacing

[IFRs]

I've been looking at API 650 table 3-6 for regular style re-pad nozzle elevations. Table 3-6 column 8 footnote d says that the HN dimensions are for Appendix A tanks only. Looking at Appendix A, section A.5.2, all you need to watch out for is 2 1/2 times the shell thickness between toes of welds. These yield different minimum nozzle elevations. For example, for a 4" nozzle, table 3-6 says 9" is the lowest a regualr type nozzle can be but if you calculate per section A.5.2 for a 1/2" shell and 1/2" bottom the number is 6 + .5 + .5 + 2.5 * .5 = 8 1/4". This number gets lower for thinner shells and bottoms. Am I possibly misinterpreting A.5.2? If I am reading A.5.2 correctly, why does table 3-6 have footnote d written as it is?

 

[TankDude]

I believe that column 8 of Table 3-6 provides the minimum centerline nozzle height for App. A tanks and is exclusive of the 2.5 times the shell thickness which is for toe-to-toe requirements between connections (i.e. not including the shell-to-bottom joint). Section A.5.2 reads that section 3.7.3 does not apply except for the 2 1/2 times the shell thickness at the connection. Section 3.7.3.3 is an additional condition to this for the shell-to-bottom joint as 3" minimum. This is how I have interpreted it in the past anyway. I typically follow Table 3-6 regardless of an App. A design or not unless there are special requirements from the client needing a lower connection. I hope this helps.

 

[IFRs]

Ah yes, but the title of table 3-6 is "Minimum dimension between weld toes or weld centerline(1)(3)". Footnote (3) says that for appendix A tanks the spacing is 2 1/2T "toe to toe of adjacent welds". This should then include the corner weld. So it would appear that if a penetration has adequate reinforcing, it can be lower for appendix A tanks.

 

[TankDude]

I can see where you could interpret it that way, however I stand by my interpretation that this does not include the corner weld as it is a special case. I am taking the conservative route. This would be a good technical inquiry to send to the API 650 commitee.