I am doing the stress analysis of a gas metering system (10” duplex line) that consists of two parallel metering trains. Each of this metering train has got a measuring instrument with block valves on either side. Both the trains are working (hot) while in operation currently. All flange connections in the system are Compact flanges.

Now the operator wants to do some maintenance on one of the measuring instrument while the other train is working. Basically what they want to do is, isolate the measuring instrument by closing the block valves on its either side and take the instrument out from the line. In order to do that they have to pull this line on one side by about 35mm to take the Seal Ring out, while the other train is working (hot).

I need to check if this external movement of 35mm on one side of the train is OK stress wise, while the other train is in operation.

I have done the stress analysis with an external movement of 35 mm in the calc. The calculated displacement stress level (by including 35mm displacement) is within the allowable. The flange loadings are OK. The nearby pipe supports are checked for the new increased loads and found they are OK.

This imposed 35mm external displacement is a one time displacement. My question is, do I have to include this while calculating the total displacement stresses?

Do I have to consider this as an occasional load case and compare with corresponding allowable (1.33 Sh)? Or compare with a more allowable, may be 90% of the yield stress.

Do I have to consider something more in the analysis?

-------
Thanks,
JDK

As usual, careful reading of the B31.3 Code would answer your question. If the expected duration of maintenance interval is below the threshhold value (10 hours for one event not to exceed 100 hours per year) you can do this. Of course, since you are aware of the 33% allowed overstress, and the paragraph where this information is presented occupies a bit less than two lines in the Code and also tells you the allowable duration threshholds, I suspect you meant to ask a different question. I will guess at that question, then answer it.

Q: How can I get around the 10 hour window given in B31.3, Paragraph 302.2.4(1)(a), and be able to consider a maintenance outage as an occasional load qualifying for an allowed 33% overstress?

A: You can't, unless you relieve the stresses in the assembly. But relieving the stresses may be easy to do for a situation such as you are describing. After you have isolated the valve train you wish to work on, break a flange pair in the train under repair. You were probably going to do this anyway, so do it within the 10 hour window. This will make it tough to bolt things back up, since you will immediately develop a 35 mm gap in the train under repair, and your maintenance crew will have to figure out how to bolt this back up (I doubt that any of your studs and bolts are 35 mm over length.) but you were also going to have this problem anyway, and it's a solvable problem (buy one set of over-length studs or bolts).

Needless to say, you will need to do an analysis case with the connectivity at one or more flanged joints to make sure that the temporary loss of connectivity in the branch being maintained doesn't throw the weight loads out of their allowable range.

How did I do?

Unless I am missing something, I read B31.3 section 302.2.4 (1) to mean the SUS condition, not the EXP condition. The 33% increase under this section relates to pressure design at perhaps elevated temperatures. I think this 35 mm displacement is either a cold pull, which doesn't affect code stress, or else can be handled as an EXP stress subject to the normal EXP allowable. I believe that even occasional displacements are handled under the expansion stress considerations. Other opinions are welcome on this.

Thanks CraigB and Loren Brown for the comments.

Can't we consider this 35mm 'displacement' as an occasional case due to the following reasons?

1) It is not a cyclic loading but it is a one-time loading. It is just one cycle.

2) In order to get this 'displacement', you need to apply a force onto the flange and this force remains there on the piping system during the maintenance of the measuring instrument. Hence for the whole maintenance time, this external force is acting on to the piping system. I would like to compare this external force with that of an additional weight for a short period of time. If we think that way it is a primary loading. Am I wrong in thinking that way?

3) If we consider this as an occasional loading, we are comparing the calculated stress with a lower allowable (1.33Sh) than that of the secondary stress allowable. In that way we are in the conservative side.

Regarding the B31.3 section 302.2.4, I think that relates to Occasional variations in Pressure and Temperature above design conditions. There I believe we should use the 10hrs for one event and not to exceed 100hrs per year.

The section 302.3.6 says if it is an occasional load case like wind, earthquake, etc we can use 1.33Sh. Here the code doesn’t say anything about the duration of the loading.


Thanks and Regards,
JDK

You can of course apply a force that will result in the 35 mm of displacement that you want. Can that be used in lieu of a displacement so that you can consider it as an occasional load case subject to the 1.33Sh allowable? I don't know, but maybe. I would be interested in hearing CraigB's interpretation on this as he has far more experience with B31.3 than I do.

I agree with Loren. The displacement is secondary and should be qualified as such. B31.3 does not give any additional allowance for occasional secondary stress.