API 653, 650 – Adding a 30” Manhole to an API Tank – Do I need an Insert?

[mearg]

Dear All,

I was asked to add a 30 in manhole to a 61,4 m API tank that has been in service since the 70s at an oil storage company in Argentina. I was specifically requested to comply with API 650 (12th edition) and 653 (5th edition). I work as a consultant and I have already delivered all the drawings and documents on how to go ahead with the modification procedure recommended by API 653 and API 650. But the company sent back a review saying my decision of having to use an insert might be incorrect. The company’s point to object my decision is the fact they haven’t used inserts in their recent modifications, but I am pretty sure API 653 proves them wrong in my particular case. I would therefore appreciate if you could help me out here to double check if my interpretation on API 650 and 653 is right.

Based on the fact that Section 9.8.6 (API 653) states that: “…Penetrations larger than 2 in. NPS shall be installed with the use of an insert plate if the shell plate thickness is greater than ½ in. and the shell plate material does not meet the current design metal temperature criteria…”, I understand the use of an insert is required, but I am confused about the “temperature criteria” part:

The lowest temperature in winter at the company’s that location is -17 C, and looking at figure 4.1.a Group I (I am using AMST A-36) on API 650, I concluded the design metal temperature criteria is not met and therefore 9.8.6 applies, which means I have to use an insert. But on the other hand, section 4.2.10.3 (API 650) states that if a plate’s design metal temperature criteria is not met, you could still use that plate if an impact test is carried out (or if 4.2.11.4 applies). My question is then: If I carried out and impact test on the metal I am going to use to manufacture the manhole and the result is positive, could I be exempt from using an insert in my design?

I am attaching a drawing of the manhole design.

Some info on the tank I am working on:
Location: Punta Alta, Argentina.
Plate Material: The company isn’t sure about it and told me to assume it is ASTM A-36
Fluid: Different types of crude
First Course height: 1500 mm
Plate Thickness First Course: 1 in.
Plate Thickness second Course: 22 mm

 

[IFRs]

Yes to your question. If the metal is resistant to brittle fracture at the design metal temperature ( not the same as the lowest one-day mean temperature ) then even though it is over 1/2" you do not have to use an insert as long as it is API Group III or lower material. However, guessing the material is A-36 is not good enough. You have to KNOW what it is - the risks are too high. They can take a coupon out where the manway will be and get it tested for chemical and mechanical properties including toughness. The pre-qualified limit for one inch thick A-36 Modified is +20 DegF [-7 DegC] and for plain A-36 is about 50 DegF [+10 DegC]. You need to find the "Lowest one-Day Mean Temperature" to determine the "Design Metal Temperature" which is the temperature the material sample should be tested at for toughness.

Looking at your drawing the round insert plate does look not large enough, see weld spacing requirements between fillet welds and butt welds in API 650.

Your path forward: Identify the material's grade, strength and toughness. Grade to determine if it is pre-qualified and so a weld procedure can be written, strength to make sure your new materials are strong enough and toughness to determine if an insert is required.

 

[thaidavid40]

A 25mm (1") fillet weld all-around? That's a LOT of weld passes for a fillet weld. Could this weld maybe be made cheaper (and better) as a single-bevel, full-thickness, groove weld?
Dave

Thaidavid

 

[JStephen]

The 1" fillet is typical there, and yes, it's a lot of welding.