API 653 floor replacement

[Mito57]

We have a 650 tank which already has been a second floor installed on top of the original following API 653. We installed a leak detection system between the two floors and lately it was noted that there was a leak. Tank was internally inspected and we are considering a new floor replacement.
Somebody put on the table the idea of installing a 3rd floor on top of the other two. I have looked at the Code and did not find anything against this if all previsions are followed. Still I don't feel 100% comfortable with the idea because I think this 3rd floor may lack proper support.

Due to the nature of the contents that have leaked between the two floors there is some concern on welding to the second floor. This is the reason why this idea was presented.

Anybody had done this before? Any feedback?

 

[IFRs]

It has been done quite a few times. Proper support of the tank is achieved by making sure the dead shells are properly aligned above each other and welded to the new bottom above. I would be more inclined to do it with a cone up bottom especially if the spaces between the bottoms were filled with sand ( which can displace under load ). I would be hesitant to do it on a tank that required an annular ring unless I could confirm uniform support under the full width of the annular ring. Have you established and eliminated the cause of the failure(s)?

 

[Mito57]

Thanks for your feedback. Tank does not have an annular ring. First floor failed due to internal corrosion. Water accumulation not properly drained. The second floor leaked after 10 years and is due to weld defects and improper installation. A few pinholes were found in welds and the overlap between plates was less than the Code required minimum distance probably causing the welds to flex and crack in a few spots. Some plates do not overlap at all.

 

[Duwe6]

Personally, I'd fix the weld defects in the second floor, using E7018 or a similarly ductile gas-shielded FCAW filler. The areas with little or no overlap can be trimmed and butt-welded [takes a Good welder, not a typical 'tankie' floor welder], or overlaid with a 2-ft wide strip.

With a water problem, if you go to a third floor put a slope on it and put in a draw-off sump. Either coat that sump or make it out of stainless steel.

 

[IFRs]

Sounds like you need a new tank contractor as well as a new tank bottom. You might consider cutting the tank just above the corner weld, removing the entire floor, removing all the contaminated sand which might pose an environmental risk given that the first bottom has holes in it, putting in a liner and new sand and then a new tank bottom.

 

[Mito57]

There is no sand. The space was left empty and a leak detection system installed. The original floor was sealed (lined) before putting the second. Still there is concern on weld repairing the second floor due to the difficulty on cleaning the space in between properly.
We are considering the repair options at this time. No decision has yet been made.

 

[IFRs]

The space between the first and second bottoms was left unfilled?

In your first post you say "We have a 650 tank which already has been a second floor installed on top of the original following API 653."

API 653 9.10.2.1.1 Suitable noncorrosive material cushion such as sand, gravel, or concrete shall be used between the old
bottom and the new bottom.

Even if the two bottoms are very close together ( 1" apart is not that uncommon ) there should be material between the steel plates.

 

[Duwe6]

Typically, between-floor "Leak Detection" involves serrated bar grating, installed tits-down. If that is your case, purge that floor's interstitial space with either argon or nitrogen, verifying proper purge at all of the leak-off valves. Without oxygen, you cannot have a fire or explosion during the repair work. NOTE: if nitrogen is used, plan on the purge taking about a day or so - N2 purges require about 10 volume changes [Ar only takes one - displaces O2 rather than diluting it]. Purge will require continuous monitoring during the cutting activities - the openings in the floor will allow O2 to enter.

Do NOT use steam for cleaning & purging, ruined a double-wall tank doing just that. 3/4" hose of 50-100# steam with only one 1/2"NPT outlet open. Inner tank walls crumpled and floor humped up. I love operators.

 

[Mito57]

That's correct. A grating was installed between floors. I thought that was clear enough when I said that there is a leak detection system.
As I said. We are considering options. Still has not been decided if it will be repaired or replaced.

 

[Duwe6]

Much cheaper to fix the relatively new floor than to replace it. And the new floor may have its own set of weld defects. NEVER seen a lap-welded floor over 15-ft diameter that didn't have at least one leak or visual reject 'pick-up' to be fixed after the initial welding. And slotting in yet another floor will probably cause you to have to raise some [most?] of your lower nozzles - not difficult, but not trivial. And that requires x-ray [or phased array].