Vessel in Cyclic Service

[nashof]

Hello All,

In my project we're now designing one vessel which, according to procces data sheet, is expected to have cyclic service conditions in its operation.
The vessel is designed by ASME VIII Div.1, but the client specification asks to study cyclic services by Div.2, which is normal.
Considering the number of cyclics and the magnitude of these, and according Div.2, a fatigue analysis shall be done.
Who is supposed to do this? Me(engineering company) or the vessel manufacturer, as part of the vessel mechanical calculations?
What is exactly a "fatigue analysis"? FEA of the vessel? FEA of the joints nozzles-shell? ¿??

The vessel is a horizontal flare KO drum, with MDMT of -170ºC, and material is SA-240 304L.
Nozzles will be self-reinforced.

Thank you very much in advance for your help!

 

[eliebl]

The manufacturer is responsible for code compliance. They are the ones applying the code stamp. In the case of cyclic service, it is beyond the scope of div 1 so by the application of U-2(g), good engineering judgement would say to use Div 2 for the cyclic service analysis. If FEA is to be employed rather than the rules specifically stated in Div 2 then it would have to be acceptable to the jurisdiction as well as the engineering company.

As a side note, the small nozzles need to be checked for reinforcing as the exemptions of UG-36(c)(3)(a) would not be applicable.

EJL

 

[nashof]

Thank you for your quick response, EJL.

The vessel will be installed in Europe, and no U stamp is required, just the CE marking for Europe, but in any case ASME design code is followed.

The vessel design will be performed following Div.1, but, as you say, the cyclic service analysis will be performed following Div.2.

We are asking the manufacturer to:
- Provide wear plates, with rounded edges, for the saddles.
- Do not use UG-36 exemptios (as you said)
- Use self-reinforced nozzles for those bigger than 2"
- Use LWN for 2" nozzles

Let me ask you another thing, please:
What is the manufacturer supposed to "study" in the fatigue analysis? Nozzles? The whole vessel?
Which could change in the vessel after this analysis? Bigger reinforced nozzleS? Nozzles not close to welds?

Thank you very much again!

 

[TGS4]

In general, I agree with eliebl. Some of my comments below are intended to add to their opinion.

Consider that the cyclic loading is, in essence, an additional loading in accordance with UG-22. How would you deal with an another load in UG-22 for which there are no rules provided in Division 1? The answer is provided in U-2(g).

in U-2(g), it says "...the Manufacturer, subject to the acceptance of the Inspector, shall provide details of design and construction which will be as safe as those provided by the rules of this Division". Usually, per U-1(a)(3), "Engineering Judgment" usually pushes the designer to follow the rules of Division 2. However, subject to the acceptance of the Inspector, that means that you must follow ALL the fatigue analysis rules of Division 2. That doesn't necessarily mean that you will be forced to use FEA, but it is typically the most convenient method of complying with the fatigue rules in Division 2.

So, in answer to your first question, the responsibility is on the Manufacturer to provide the calculations. However, it is your (the user or user's agent) that has the responsibility to define the cycles (be they thermal, pressure, or otherwise).

As far as the scope of the fatigue analysis is concerned, if it were me, I would want EVERY aspect of the vessel considered. Note that I didn't say analyzed, only considered. How do you know, a priori, what the weakest location of the vessel is going to be, vis a vis fatigue? Unless you've been doing this for many, many years, you probably don't.

Yes, you're going to want to use common sense (or is that good Engineering Judgment?) in regards to construction details. I would not permit the use of fillet welds, particularly on reinforcing pads or at the vessel support(s).

One final note - I would HIGHLY recommend that the Manufacturer be directed to use the rules in the 2007 (or 2008) Edition of Division 2. The fatigue rules in the newly-rewritten Division are significantly easier to understand and implement, particularly the fatigue of weldments issue. In the end, I suspect that the fabricator will not have the experience necessary to perform the fatigue analysis themselves. Again, highly recommend that they hire someone who has that experience. You'll pay a non-trivial sum for the engineering work, but it will be worth it when your vessel doesn't fail after only a few years of service.

 

[vesselfab]

thank you for your opinion about fabricators.

maybe some....but not all

some of us even have computers now

goooolllly

 

[TGS4]

vesselfab - didn't mean to paint ALL fabricators with the same brush. Based on my experience, most (but you are correct - not all) are not able to perform fatigue analyses.

 

[eliebl]

Given the climate where I work and live, and I think vesselfab would agree, there may not be many at the engineering houses that could perform these calculations.

Choose wisely when you farm this portion out to someone.

Having an MDMT of -170 C is lower than I have seen on KO drums. Is there going to be significant thermal cycles as well as pressure? This may almost be an interesting design do.

EJL

 

[TGS4]

If, by engineering houses, you mean EPC companies, then I would go further to say that there very, very few out there that could do this level of work. However, there are many consulting companies for whom this is their bread and butter, and do this type of work every day.

 

[nashof]

Thank you very much for all the responses!

The MDMT is -170ºC, which corresponds to LNG going into the vessel, once the LNG is in the vessel we have an electrical heater in lower outside part of the vessel which gets 200ºC and heats the LNG to a temperature of aprox. 100 ºC, and then it goes to the flare.
(I'm not a process engineer, but more or less this is the process..).
The number of cyclic (our process department has defined this) is 360, and for the magnitude of the cyclics I'm not sure if we should consider -170 to 200 or -170 for 100.. ¿?

And finally, I would like to say that I work for a big EPC company and we are supposed to have the media and possibility to perfomr this FEA, but honestly I don't think we have enough experience with this.

 

[yl0525]

Hi Nashof,

Allow me to add my little opinion here. Fatigue analysis is mainly to study the fatigue life of a product, ie vessel for your case. Since the heater is installed outside of the vessel which is in fact to heat up LNG to approx. 100ºC, you should study on the highest possible wall temperature (perhaps include some tolerance on top of it) for the vessel but not the temperature of the heater.

As for my company, we always sub-contract out FEA if it is in our scope of supply and shall be guarantteed, because we do not have experience and capability to perform this analysis.

 

[nashof]

Thank you for your opinon yl0525!
My doubt about wich temperatures should we study was because in the inner wall of the vessel temperature will change from -170 to 100, and in the ouside wall it will change from 200 to ... (I'm not sure what temperature).
My company, as you said, will not perform the FEA, but we are supposed to define number of cyclics and magnitude of them, defining the "F" number, according ASME VIII Div.2.

I think we should consider temperature changes between -170 and +100.
With this 270ºC difference, the F number in Div.2 and with the number of cyclis we have(360), we'll have more than "1000 cyclics", as defined in Div.2 Ed.2007, and so a fatigue analysis must be performed.

The manufacturer will be responsible for doing this, and they could perform a FEA or comply strictly with all the calcs in Div.2, submitting the calculations to the EPC company for revision/approval, as they do for mechanical calculations, drawings, etc.

Which could be the consequences of the fatigue analysis?
I mean, what changes could the vessel need? Bigger reinforces for self-reinforced nozzles? Differente geometry in welds?