Old topic, but i would like to have your opinions

(We) used to send out a nozzle load matrix for all vessel manufacturers to comply with when ordering vessels.
IE, your nozzles will accept such loads as: fx fy fz mx my mz blah blah blah.

I've never liked this as:
1.we are telling the vendor what to do.
2.Cost implications are huge.
3.Stress dept never gets to see the vessel drawings before its ordered!

I'm more in favour of the vendor telling us what the nozzles can withstand and reviewing results to see if this is acceptable. in particular, ensuring that the nozzle for the emergency relief device is strong enough. I feel that nozzles should only be reinforced if there is a particular requirement.

So, what happens at your companies?? i will be intrested to know

SUPERPIPER:
When we send out RFQ's for vessels and other like equipment, we request that the vendor either provide the allowable nozzle thermal loads or approve the buyer provided nozzle thermal loads resulting from the stress analysis. Typically the fabricators will not provide any loads, but will review and approve the buyer provided loads. Sometimes they will reinforce nozzles accordingly, or we have to provide more flexibilty to get approval.

Having sat on both sides of the fence on this one, my thoughts are that it is best for the buyer to specify required loads to the vendor rather than the other way round. Here's why.

1. The buyer can set the allowable loads to what he sees as reasonable. That way the vendor should include for these from the RFQ stage onwards and the piping stress team can work away before the vessel is truly designed, relatively safe in the knowledge that if they keep their imposed loads below the allowables everyone should have an easy life. If the allowables are way too high, the vendor will soon tell you.

2. If you rely on vendor imposed allowable loads, it is in the vendor's interests to make these as low 'as he can get away with' largely because this makes his life easy. Chances are that because whoever is specifying the vessel has not thought about nozzle loads [obvious - because he hasn't asked for them], any vendor provided figures will be glanced at and regarded as acceptable. Later on in the project when the stress team cannot meet these loads, the vendor is well within his rights to go for an 'extra' to cover the engineering time and costs in meeting the actual required loads. This can be a pretty substantial figure depending on how late in the procurment process this arises.

The big question to me always to be aware of is where the allowable nozzle loads are intended to be taken. It is common to believe that because the loads are 'nozzle' loads, that they are taken at the flange face. Sometimes [like pumps] this is true. It is the convention however in vessels [PD5500 / EN 13445 / WRC 107] to take the loads as acting at the nozzle/shell interface as this is the critical point. If your loads are specified as acting around the flange face MAKE SURE THE VENDOR IS TOLD THIS EXPLICITLY!! If you consider a bottom outlet nozzle on a vertical vessel that comes off the domed end and then turns to exit horizontally through the skirt, the moment arm between the flange face and the nozzle/shell interface could be very large indeed. Imposing large forces on the end of this moment arm could either lead to a well over-engineered and very expensive vessel or even end up in court as a contractual matter [I know of at least one of these currently on-going between a vendor and a contractor - I reckon the vendor is right and the contractor is being an eejit IMHO]

As far as the vendor giving absolute figures for each nozzle I think this is too impractical. The degree of interactions are too complex to be easily tabulated. Not only does the position of the nozzle in relation to a stiffener [ie head] affect the maximum allowable loads, but there is in interaction between pressure, radial force and longitudinal and circumferential moments. EN 13445/ PD5500 App G2.8 defines this interaction as:-

[[max (P+Z), (Z), (P-0.2Z)]^2 +B^2]^0.5 <= 1

where
P is ratio of design pressure /shell maximum allowable pressure
Z is ratio of radial force / maximum allowable radial force
B is Root of the sum of the squares of the ratios of circumferential and longitudinal moments / their respective maximum allowable moments

[hope that is clear] So any decrease in radial force can be accompanied by an increase in the moments applied, any increase in wall thickness local to the nozzle [ie at cone/shell junction] allows an increases in applied loads. Note also that the nozzle torsion and shears are NOT regarded, so potentially these could be pretty large without any impact on the vessel integrity.

Giving the vendor your required loads in tabular form may not be perfect, but it is pretty simple for everyone to work to.

Gentlemen... keep going this is a great topic and I enjoy reading the various opinions before I weigh in with my (quasi-worthless) opinion.....

Great replies cheers,

Now then, the problem is what magnitude of loads do we give to the vendor?????

there are so many different types of vessel/tanks/exchangers/ etc. how can a chart attempt to cover all the different designs?

It is our experiance that vendors will design to the loads given, resulting in hugely overcosted designs.

how do you communicate that a nozzle is carrying a level transmitter and does not need robust design?


after reading the replies, maybe the way to go is to issue a force matrix, but request that any extra cost or design required to meet the matrix is approved by the purchaser first.

this though requires thought and procedure and time, rare commodities in todays age.

There are many types of vessel/tanks/exchangers/ etc, but many follow standard forms eg cylinders, dished ends etc, where there are a few trusted methods of calculating the local stresses and flexibilities at the critical point [whatever that may be]. The vessel codes + WRC covers most things with backup from the likes of API 650 for large diameter tanks. For weird stuff like plate exchangers you are going to have to speak to the vendor as the designs are very proprietory. You could however suggest using say 50-60% of your 'normal' allowables as a starting point. Not very scientific I accept, but it is at least a starting point for both parties.
As for determining the values themselves.
I have seen a couple of rational methods based on nozzle sizes, but the results are generally ok for most purposes. Very large diameter nozzles [relative to shell] allways require special treatment as identified by the codes.

One typical [contractor] method is :-
Fs = Force Allowable Stress MPa
Ms = Moment Allowable Stress MPa

Values of Fs and Ms for each pipe size:
Pipe size [in] Fs Ms
<=6 2.1 41.4
8 2.1 39.1
10 2.1 36.8
12 2.1 34.5
etc etc

Am = metal area [mm2]
Z = section modulus [mm3]

Allowable Resultant Force FR = Fs x Am [N]
Allowable Axial Force FA = 0.5 x Fr [N]
Allowable Transverse Forces FB and FC = 0.6124 x Fr [N]

Allowable Resultant Moment MR = Ms X Z X 10-3 [Nm]
Allowable Axial Moment MA = 0.7071 X MR [Nm]
Allowable Bending Moment MB and Mc = 0.5 MR [Nm]

One typical [vendor] method is :-
Cylinders FB=K.2000.D
FC=K.1500.D
FA=K.2000.D

MB=K.100.D²
MC=K.130.D²
MA=K.150.D²

Spheres FS=K.2500.D
FA=K.2000.D

K= 0.6 FOR 150# 0.7 FOR 300# 0.8 FOR 900# etc

Regarding instrument nozzles, it is all very well saying they see no load, but what will happen in 5 years time when the owners want to revamp the vessel and use the instrument nozzle to fit some process line. Much easier to make them all the same. The possible exception to this is perhaps defined manholes.
[Phew! it's been a day of long posts.............]

WOW
Fantastic
Thanks Kenny,
I'll print it all off and mull it over a (beer!!)drink later on.

"....hugely overcosted designs..."

How much does a repad cost?

For shell & tube exchanger and pressure vessels, we issue a standard force/moment based on nozzle size as part of the inquiry. I don't remember the exact basis for them, but we are generally able to meet them w/o a lot of weird routings and it's a rare thing to have the vendor come back and say they can't meet them.

Quite frankly, we generally can't wait for firm vendor data to do stress analysis. The vessel/exchanger engineers issue a basic outline drawing showing nozzle locations and orientations to the vendor. Piping design starts building the piping model from that drawing and we start working those ISO's.

Maybe in the "good old days" when you could spend a couple of year engineering a job, you could wait for all the equipment data to be issued before piping and stress got in deep, but today's "fast track" schedules are less tolerant of such things.

This question was studied before many of the current participants were born. Drs. Markl and Rossheim felt that the weaker the nozzle the greater the cost would be in making the piping system adequately flexible to meet the lower nozzle load.


So they suggested buying static equipment with the following minimum strength in the nozzles…

The breakout of the Markl Rossheim limit as follows
D = Pipe Outside Diameter in Inches + 3 Inches

Radial or Punching loads…. 3.25 D^3 Lbs F
Tangential Shear in any Direction…1.5 D^3 Lbs F
All bending moment… 60D^3 Ft Lbs M

This can all be found in piping Systems Design by M.W. Kellogg company an out of print book that is hard to find and expensive.

Anyhow as pointed out by Mr. Klein repads are quite inexpensive usually and are quite helpful at increasing the nozzles strength. Some unenlightened owners or consultants simply order their vessels with no allowance made for mechanical loads. Having worked on these projects I can tell you the money saved on a cheap repad was more than burnt up meeting excessively low target numbers.

Captain Kenny, do you have a html link for the formula you have wrote with vivid explanations or the reference..

Which one?
The load combination formula is from PD5500 ApG2.8 or EN 13445.
The formulas for allowables laods are from proprietry documents which I cannot send out.
which bit don't you follow?

Just got back from vacation or I’d have chimed in on this one earlier.

I’m in agreement with Cpt. Kenny et al here. Much better to let the vendor know up front what loads you want the equipment to withstand. Recently I was involved late on a project in which the loads had been left to the vendors’ discretion. I encountered two time-consuming and frustrating problems. The first was that I had difficulty getting allowable load information from the vendors. In some cases the vendors demanded payment for the additional calculations involved since it was not part of the original contract. In other cases the loads were received long after the stress analyses were completed. The project is now in construction and some of the allowables have still not been received. The second problem was that vendors provided allowables that were difficult or impossible to meet. In one very lengthy email exchange a vendor repeatedly said that his nozzles were not designed to accept any stresses. When I finally protested that even bolting on an instrument would add some stress to the nozzle the vendor replied, “The vessel nozzles are not designed to accept piping loads. All external piping should be independently supported.” Exasperated, I sent my CAESAR II calcs to our in-house vessel engineer and told him to take care of it.

For many years I’ve asked the vessel group to include criteria in the RFQ essentially identical to Mr. Luf’s above. I picked these up from some long-forgotten source in the 1970’s. (I confess to being unaware of their Markl-Rossheim origins.) Our proposals now include the statement:

Vessel nozzles shall be designed to accept the following loads:

Axial force (lb f) = 3.25(D+3)^3
Lateral force in any direction (lb f) = 1.5(D+3)^3
Bending or torsional moment (ft lb) = (60(D+3)^3)/12

Where D = outside diameter of the nozzle pipe size (in).

I should add that Kenny makes an excellent point concerning the nozzle/shell location of the load. I shall make sure that’s clarified in future proposals.

Finally, I offer a lesson-learned from my personal school-of-hard-knocks inventory: Never, never, never purchase an FRP vessel without an up-front agreement on nozzle allowables. FRP vendors seem always to design for zero loads. I am frequently reminded of a case in which the vendor successfully quadrupled the cost of vessel in order to redesign for very reasonable piping loads that had not been included in the original RFQ.