HYDRO PRESSURE - VERTICAL VESSEL (2) HEAD THICKNESSES

[tr6]

We have a vacuum tower in construction - 8 M dia x 30 M S/S. Designed for 8 barg. The top head is thinner than the bottom head. The vessel is vertical when in operation.

Due to the size, it is not possible to hydro the vessel in the vertical position and it will have to be hydro'd horizontally supported by saddles. Calculations have been performed to establish the number and position of the saddles.

Question: What should the hydro pressure be so that both the top head and bottom head are adequately tested without overstressing?

I'm concerned that if you base the hydro pressure on the top head, the bottom head is not adequately tested. If you base the hydro pressure on the bottom head, the top head will be over stressed.

 

[GenB]

Interesting point, you may be in big trouble. The test will be on the weakest part. Loadings has to be considered by Code in this case the head pressure.
I will consult my AI to see what he has say.
ER

 

[prex]

As already pointed out by generalbir, the test pressure is to be determined as the minimum of the allowable test pressures for different parts of the vessel, and you will satisfy the code requirements even if the lower head is not fully tested.
However the code defines a minimum test pressure, not a maximum one. Hence you may determine the test pressure based on the limits set forth by Div.2: general membrane stress not higher than 90% of the specified minimum yield stress at test temperature (and primary bending stress not higher then 1.35Y, but this shouldn't be applicable for you). The resulting value could solve your problem, depending on the design temperature and material type of your vessel.

prex

http://www.xcalcs.com

Online tools for structural design

 

[arto]

We usta test smaller diameter towers on their sides to avoid head pressure differential, but on an 8-m diameter, you'd probably need to brace it.

 

[JoeTank]

tr6,
You are getting yourself wrapped around the axle for no reason. ASME VIII requires that all portions of the vessel be subjected to a pressure that 1.3 x MAWP. The MAWP is what is stamped on the nameplate. Lay the vessel on its side (properly braced of course), fill it up with water and pump it up to the test pressure. Yes, the lower head will not see it's 1.3 x the pressure it will see in operation, but that's what just the way it is. You have met the Code requirements.

Steve Braune
Tank Industry Consultants

http://www.tankindustry.com

 

[jte]

tr6-

Take a look at UG-99(c) and Appendix 3-2 under "calculated test pressure." Basically, this is saying that you may test based on 1.3*MAP where MAP is the Maximum Allowable Pressure, new and cold. Since you probably have some corrosion allowance, this will effectively raise your hydrotest pressure without worrying about overstressing the top head.

You didn't state which type of heads you have; be careful to check the compressive stresses in 2:1 SE heads and conical transitions if you exceed 1.3*MAP. Remember that 2:1 heads will fail through wrinkling near the knuckle due to the compressive stresses at that point when under internal pressure.

jt

 

[JoeTank]

jte,
You are not required to calulate the test pressure when you choose to test at 1.3 MAWP. ASME permits the use of the design pressure P being set equal to the MAWP for the nameplate and testing. Regarding temperature, you must adjust the test pressure by the ratio of the allowable stress cold divided by the allowable stress hot.

P test = 1.3 x MAWP x S(cold) / S (hot)

Steve Braune
Tank Industry Consultants

http://www.tankindustry.com

 

[unclesyd]

Why wouldn't the top head be designed/evaluated for the Design Pressure of the vessel and the bottom head/shell for the same pressure + additional considerations for the water/process liquid column, flooded conditions. If the support is causing the problems of testing the vessel in the vertical or installed position it should still be tested after installation. Flooded conditions have to be considered in the support design. Horizontal testing takes these considerations out of the picture.

We had numerous vacuum columns (12 M dia. x 25/30 M) designed for F.V. and 50 PSI and all were tested vertical and full of water. All these columns had thicker bottom heads and the shell courses decreased in thickness for each course. All these vessels were tested in the vertical position either by measuring pressure at the vent or at the bottom after correcting for the liquid head. This testing procedure continues throughout the life of the vessel.

 

[jte]

The adjustment for temperature goes without saying.

My impression from tr6's question was that he was interested not in meeting Code minimums, but in exceeding them. Many companies have requirements to exceed the Code minimums for hydrotest. It happens often enough that the basis for the optional hydrotest pressure is spelled out in the Code as I pointed out in my previous post and it also has a nickname: MAP or maximum allowable pressure (new and cold). Some commercial PV design programs (eg: PV Elite) will automatically calculate both MAWP and MAP.

Using MAP as the basis for the test pressure becomes significant when you have a corrosion allowance which is a significant part of the wall thickness. If the philosophical idea of a hydrotest is to stress the vessel beyond any anticipated level of stress it could see in the field then you must use MAP. If you merely want to meet Code, feel free to set the MAWP at the design pressure (leave that extra corrosion allowance or pressure on the table as wasted), and use a hydrotest pressure calculated as 1.3(DP)(stress ratio).

Let's say I have a vessel with radius R, new wall thickness 3/8" thick with 1/8" CA operating at say 200° so the stress ratio is 1. The stress at Code minimum hydrotest will be
Smin = 1.3*MAWP*R/.375 = 3.467*MAWP*R

However, the MAP will be (3/2)*MAWP so the MAP based hydrotest stress will be
Smap = 1.3*1.5*MAWP*R/.375 = 5.2*MAWP*R

The stress while in operation at MAWP at the end of life of the vessel (fully corroded) will be

Seol = MAWP*R/0.25 = 4*MAWP*R

In this case, the vessel shell will see more stress in operation than it will in the shop hydro based on MAWP. Using the MAP as a basis ensures that the vessel sees more stress in the shop than it ever will in the field.

The situation which tr6 is dealing with can easily make the bottom head/shell see a lot less stress in the shop than they would in the field. That's the incentive for not merely meeting Code minimum requirements.

jt

 

[JoeTank]

jt,

Perhaps I have read the original post to hastily. I thought that tr6 was worried about meeting the Code minimums only.

When I still had hair and did my ASME calcs by hand, we typically used the design pressure in the owner's spec as the MAWP (as permitted by ASME VIII). I have come to understand that this was typical of most vessel fab shops. Only when required by the client did we do the MAP calcs for the vessel in the as-designed condition and then base the hydrotest on the calculated result. Today this would much easier as many shops use those new slick programs.

Oh by the way, ASME VIII only states minimum test pressures. Strangely, it is silent on what happens if you want to use a higher test pressure. There is a common folk lore out there that the hydrotest stresses must be limited to 0.9 of yield. It is not in ASME VIII. I have always considerd 0.9 of yield as the limit, but it's not in the Code.



Steve Braune
Tank Industry Consultants

http://www.tankindustry.com

 

[GenB]

After going through all the posts, I think I have a position
only if you have more metal than needed on the upper parts of the vessel. you calculate to a higher pressure and test at a higher pressure to meet the bottom head expected operating pressure(say the MAWP of the vessel + the head pressure),
then,
On the data report remarks: vessel designed and tested for 1xxxMAWP but stamped for 1xx MAWP because of ___________
This will meet all your demands and the Code.
Regards,
ER

 

[prex]

SteveBraune,
please review AD-151.1 of ASME VIII Div.2

prex

http://www.xcalcs.com

Online tools for structural design