LPG Tank - Shell Thickness Calculation

Pressure relief valve must be set at not more than vessel design pressure (not burst pressure) and max 10% overpressure allowed (at max relieving flow). Possibly PRV set pressure can be maximum allowable working pressure (MAWP) which could be higher than design pressure, but unlikely anybody will bother to calculate MAWP.

Seek the help of a local, certified mechanical engineer.

Someone will have to take ultimate responsibility for the design calculations.

An anonymous forum is not the place for certified calculations.

This is what could happen if the calculations are wrong:

Train car explosion bleve - YouTube

Unless the local code is crazy, you are correct, and your colleague is incorrect.

No safety factor?

The safety factor will be used later, using the formula for hoop stress and is connected to the TS and YS of the selected material. But before this I have to know the correct pressure that is acting on the shell and that will be used in the later calculation.

All of you can freely write your opinion about my question, you will not be responsable for any explosion as explained earlier. Don't warry :)

Good for you.

Glad to hear that you are using the correct design process.

I really wasn't too worried, since I assumed that all the correct procedures would be followed.

It's just good to hear it from you.

Seems our Chicken $hit rater is at work again, huh? You brought up a legitimate concern, so I took upon myself to offset the OT.

Dan

The OP's question wasn't about safety factor; it was about which pressure to use--absolute vs. differential or gauge. ASME uses the latter. Joint efficiency hasn't yet been asked about, either.

Thank you for your replys.

All the calculation will be at the end reviewed by the side of one of the classification societies. So there is not any danger that enything will be produced and delivered just like that :)

The calculation I have to do is because we need to see the approximate cost of the tank production after what the decision wheteher we will enter in this project or not will be taken.

The tank shall be used for marine purpose and built according ASME VIII D2 and IGC code.

If not already included, you may need also to account for the pressure from the height of the liquid.

Yes it is included. I wrote above that dynamic and static pressure is total 0,72 bar.

I wander why he insist that the 18,72 bar should be used?

Maybe he confused the tank filled with LPG with the tank filled with water? In case the water is in the tank then you have the dynamic+static pressure, vapor pressure on higher temperatures, atmospheric pressure inside the tank and atmospheric pressure outside the tank. In this case the 18,72 bar should be used for the shell stress calculation. But when the tank is filled with the LPG then is only LPG inside. As far as I know when you fill the LPG tank you have to first substitute the air with inert gas and then the inert gas with LPG.

If your pressure are "absolute" you can make the difference if they are "relative" they are measures against atmosphere so that the "internal" pressure has to be considered since it is already by measure the difference between absolute internal pressure and ambient.

I have the feeling that "barg" means gauge value you should check if the gauge did the measure against vacuum or against ambient.

I think that I see where I'm wrong.

Propylene on 43.8°C is 18 bar, but bar absolute? I was using it as bar gauge.

0.72 is static and dynamic calculated together. Only static is 0.2 if I remember well.

OK, I may be missing something, but I don't see how dynamic loss can increase the pressure in the vessel above vapour pressure. If you give a bit more detail, it might be possible to get rid of the extra 0.52 bar for vessel design.

Not that 0.52 bar additional safety isn't a good thing!

The stress on the shell is caused not only from the vapor pressure but also from the fluid weight (static pressure) on the lower part of the tank and also from the dynamic pressure (cargo moving in the tank during the rough sea).

When the cargo moves the impact on the shell can be significant, so it has to be calculated for the worst case and added to the vapor pressure and static pressure.

OK thanks I get it now. I had assumed your dynamic pressure was based on pipe losses.

Because of the dynamic loads in case of sea transport I would considut NOT ONLY the pressure as a criterion but also the reservoir stability as a whole. Similar problems occur for instance in rocket reservoirs with a low pressure but a lot of dynamic loads.

POGO!! This is a very good point, glad you brought it up. While I don't think the POGO levels would ever reach the frequencies and levels of a "no go" situation for a rocket, I really have no idea. Not really a POGO effect either, but it's a close enough analogue. Dynamics involved in naval transport of cargo will depend on pretty much EVERYTHING you could think of, from ship size, speed, roll and yaw, position of cargo in or on the ship, tie down practices, those are just the obvious.

So do the maritime standards have a reference guide or guidelines for all this stuff so you can design the vessel to basic dynamic loads and target resonant frequencies to avoid? Or is all of this stuff built in to the standards for vessel support?

Of course that whenever you design something that will be used on ship you have to include a lot of coefficients concerning the max waves, area of use (no limit, inland water, etc), about ship dimensions etc... It is not just like that, you have a lot of calculation to do.

It's been quite awhile, so I'm not up on the codes, but I'd say that you can't go wrong designing based on the absolute pressure in the tank, and in fact I would think that you wouldn't want to. That's just me thinking logically, if the code says otherwise or I'm way off base let me know.

For simplicity let's assume you're designing to the yield stress. At sea level you would have a positive safety factor, and if you ever launched this thing into space the safety factor would go to zero. If you subtract a bar, your safety factor would be zero at sea level, and go negative with any increase in altitude ASL.

Totally agree with some others. If you design for a 3 bar environment, it always needs to be in that environment. A relief valve set to actuate at a gage pressure of 15 bar would be required for an event which cause the environment to change.

Anyway, I haven't performed design like this since college around ten years ago. So the assumptions are liberal and I just ran things through with some pretty simple logic (no sleep, trying not to think too hard lol). So real world conditions in the tank, safety factors, etc, really weren't given much thought. It's just basic thinking on my part as to why absolute should be used.

Like I said, If I'm out in right field or the code states differently, I'd appreciate it if someone more knowledgeable could correct me. And I know you would never design to yield stress :). Always looking to learn!

Steve

Just a comment on safety factor - an appropriate design pressure is determined by everybody involved (including standards) and likely to have some safety. This is then used in vessel design, and the design stress allowed by the code (based on material selection, temperature etc) includes a safety factor on the yield stress and the UTS.

Hi, Lupezan: Please try specification, http://en.wikipedia.org/wiki/EN_13445.

If these tanks are to be filled to 100% of their volume in liquid. Then the vapor pressure does not work. Stress do to thermal expansion of the fluid will exceed the stress of working pressure of the vapor. Most these tanks have working pressure of 240psi(16.54bar). Tested to 1.66666 of the working pressure. And have a relief valve not to exceed 5/4 the working pressure

Because of this problem a few years back they regulated tanks to 80% fill here.

Since you have not stated the application. That extra 20% may have some value you need to look at.

Yes, later I stated the application, it is for marine purposes. According the IGC code the tank can bi max 98% full in order to always have the vapor phase.