Piping Wall thickness calculation / Axial Stress

[Achman]

Hi All,
I have been challenged by a colleague about the thickness to be specified for a piping material class for a process plant and would appreciate any input about this.
The way I have always been doing it has been to specify the thickness using the relevant code (ASME B31.3)based on the flange rating pressure, taking into account the manufacturing tolerances, corrosion allowance, etc. and then to chose the next available size up from the commercially available thicknesses, however, there are suggestions that some extra thickness should be added to the calculated thickness because of the loads that the piping systems will be exposed to during operation.
My understanding is that because the thickness is calculated based on the hoop stress and the longitudinal stresses due to pressure are half of the amount of the hoop stress, there is some room for extra longitudinal stress due to other loadings.
Besides, the thickness calculation is based on the flange rating pressure which is normally higher than the actual design pressure of the lines and also we usually end up adding 1-2 millimeter thickness when specifying a standard Sch. which means having some extra thickness available for other loadings.
So the question is : Do you guys add any extra thickness for the potential longitudinal stresses when you calculate and specify pipe thickness based on ASME B 31.3 for a process plant ?
Any comments would be highly appreciated.
Regards,

 

[SNORGY]

Not typically.

If your longitudinal stresses are such that you need to jump a wall thickness to meet code, then to me, the solution is in increasing the flexibility of the piping system, not increasing the wall thickness.

But that's just me.

 

[LittleInch]

Interesting.

Normally I would expect the piping design initially to be based on the system design pressure, which could be considerably lower than the maximum flange rating + corrosion and under tolerance then next std wall thickness. Basing pipe design on the flange rating is really quite conservative (expensive) for a decent sized plant.

Then the design is checked for stress and if you need more thickness for some other loads if you can't reduce them by making it more flexible then you add it as required, not as a standard amount.

Because loads other than hoop stress vary widely from virtually nothing to very significant amounts, it is not normally economic to simply add more metal when none is required, or maybe not enough. That's why design engineers exist.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[BigInch]

Snorgy, that is unquestionably the best approach, however if the system is a continuously buried pipeline where the line is "fully restrained" due to soil weight and friction forces, any attempt to increase the flexibility, say by reducing burial depth, or insert bends, may result in uplift buckling or overstressing the bend, respectively. Sometimes the only practical alternative is to add extra wall thickness, not to decrease the thermal stress as in the above case thermal stress is independent of wall thickness, but to decrease buckling length or to reduce the hoop stress just that extra little bit where the combined stress will pass the check.

you must get smarter than the software you're using.

 

[SNORGY]

Agreed BigInch, however, I didn't go down that path in my post since I inferred from the OP that B31.3 was the design code - most of the piping under the scope of which I seldom bury.

For pipeline work, I have done some iterative gyrations (recall my Rule Of Thumb about detailed stress analysis being required when [SMYS (MPa)]/6 <= [(T-hot - T-install) C]) when, depending on a lot of other things, sometimes the problem is solved by adding extra wall, whereas sometimes it isn't.

Your comment about "that's why design engineers exist" is actually profoundly true.

 

[LittleInch]

I think you'll find that's my comment... ;-)

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[XL83NL]

Having your spec flange limited is a good approach Id say - though somewhat conservative - youll find that you dont need to redesign everything when an increase in design conditions is required.

The only reason I'd add some wall thickness (on top of that) is for small bore piping (up to 2"), when designed for low pressure classes. When such specifications require at minimum the use of 10S (and youre using SS316 or another stainless steel), I'd use 40S for that size range.
I dont want the whole piping system to plastically deform when big fat Joe the mechanic steps onto a spool, and uses it as a ladder, so he can reach some cable way or something else.

Having pipe spans for 1 or 2" pipe of wt 2.8 mm is nothing. Replacing a spool is a lot of work. The price difference between 1" 10S and 40S is minimal.