Does anyone have any thoughts on using CAESAR II to calculate the amount of movement that is permissible during proposed repairs and/or inspections?

Subject problem:
=====================
Our facility completes regular inspections of our piping systems. And some of the old installations do not use shoes or pads at the pipe support locations.
It is common for failures to occur at pipe supports, due to external corrosion.
The inspection department is unable to access the pipe at the support, and requests the maint. department to "lift" the pipe to provide access.

There are also requests to perform the lifting procedure while the pipe system is in-service.

API Recommended Practice, 2200, Repairing Crude Oil... Product Pipelines
Section 6.2:
c. Calculations should be performed to determine the amount of pipe movement that is permissible during the proposed repair. These calculations should consider the pipe metallurgy so that proper pipeleine support can be provided during the repair.

It appears to be a bit of a catch-22.

leftofanswer,

I can see no problem in using Caesar II to calculate stresses in the piping system as you describe. Will the system be out of service and at ambient temperature? Will the piping be empty? Will there be any discontinuities caused by removing components (e.g., valves). If it is a hot pipe that will be lifted at ambient temperature it may have acquired some "self springing" and if there are strain sensitive equipment attached you may find that loads on the equipment are more limiting than the stresses in the pipe.

It will be a "one-time-loading" so it will not need to be addressed by fatigue analyses.

Wear your hard hat and hearing protection ;-)

Regards, John.

thanks john.

you have mentioned the exact concerns i have.
the lifting proposal (or request) is for various pipe service. no flange connections are broken, no components are removed.
the pipe is not empty. the pipe system is in service. (with operating pressure, temperature, and flow rates.)


in general, the service is less than 400 degrees F, and less than 150 psi.
most is CS, A-106B.
design calculations for the non-lifted pipe are OK per B31 code. but no evaluation has been completed for "lifted conditions".

i have begun to evaluate specific cases. and have chose to use the corroded condition in the caesar model. (takes 0.125" off pipe wall) but this does not take any consideration for the exact condition we are inspecting for... local corrosion at the pipe support.

special attention is given to the boundary conditions and the connected equipment. i will attempt to identify "affected zone" of the lift.
we will collect NDE (UT thickness measurents) for the section of pipe that will have increased stress due to the lift.

so there are two problems:
1.) how much should be modelled? e.g. branch lines, distance away from lift.
2.) what should be used for the corrdoed condition?

isn't anyone else faced with this type of problem?

i believe most maintenance requires lifting, moving or otherwise applying external forces on pipe.

Lefto,

You are not alone. Lifting piping systems for maintenance is done routinely in refineries. MOST of the time the systems are taken out of service because owners want ZERO (or less) risk of fire or other adventures. There is no "pat answer" to how much piping must be modeled. It depends upon how much piping will be affected down the line by local lifting and this is probably a function of the stiffness of the pipe (pipe size and wall thickness).

I would be loath to risk a lift with the possibility of a rupture spilling 400 degree F. product. I would not do it at all if the product were on the OSHA flammable or hazardous list - too risky. I would seriously weigh the risks if the contained product and the pipe metal was at ambient temperature.

You have thought of some of the "sticky" details like local wall thinning at supports. I have seen (witnessed) pipe failures on lifting. Example - NPS 14, Schedule 40 CS off-site piping was supported on concrete sleepers with embedded steel "wear bars" that resulted in "point loads". There was local crevice corrosion exacerbated by sliding (thermal movement) wear and the owner decided to lift the pipe and weld "doubler plates" at all the sleepers to extend the life of the system. Lift it, weld on the doubler, drop it back down on the sleeper. Until on one lift the pipe failed at a local line of corrosion and they had a culvert full of gasoline. Not much fun that.

The analysis will give you some feel for the stresses at the locally distresses area if you model it correctly - the thin wall (corroded) areas will be the "fuses". Also, do not neglect to consider what is happening to the bolting at the flanged connections when the additional large moment is introduced there.

Obviously the more "lift and support" locations the better. Analysis (with local thinning modeled) should give you some idea of how much you can lift any location without overstressing the areas that are not lifted. I have to be honest, I have a bad feeling about this!

John.

I boldly continue this thread with a related question.

Regarding forced displacements, what stress limit should be applied? I got a query regarding a problem on a platform where settlement had caused two pipelines (one running above the other and crossing it) two come closer then desirable. A question from a young pipe designer was if one of the pipes could be forced to displace by modifying some of the supports.

What stress limit would be applied in such a case? And also, if the displacement were only temporary?

Jon