Pipe support during excavation of gasline 2

[Dario2002]

Searching this forum shows quite many threads regarding pipe support. Most threads with specific answers are quite old (year <2005), and some general info is that there is no fixed rule or standard for pipe support distance.

Has this changed in last few years? Is there now any quide/rules/tables/calculations for maximum unsupported pipe section?

We have upcoming maintenance on long sections of natural high pressure gas lines (12", 20" ANSI 300, 7 mm thickness). Long sections must be recoted (sections of up to 300m in length). Is there any simple "rule of thumb" what length can be excavated without temporary supports?

Sure I can calculate bending stress, but there is also mater of inside pressure (around 40 bar, MAOP=50 bar) and welding which I'm not sure how to calculate and consider them or not..

What is your experience in such situations. Is unsupported section in range of 5m, 15 m, more?
Thank You for all inputs.

 

[LittleInch]

If you're doing this work while the pipe is in service then be very careful. You should be looking at a combined stress calculation, either equation based or using a stress program such as Ceasar with "soft" shoulders for the excavated section. The maximum allowable is then fixed by your design code. If your pipe is in anyway corroded or you have doubts about any of the welded joints, add a considerable safety margin.

My initial guess would be that the 12" would withstand about 10m span without getting close to your overall stress limit and the 20" probably about 15m. This is just my starter distance - you need to work it out.

If you do excavate longer sections and use intermediate supports, try and make sure that they are not point loads and that the support is at least 1m wide and "soft" at the edges. These will probably need to be closer together than for single excavations.

I've seen this sort of thing done a few ways - long sections with lots of supports or slings or excavate individual sections, re-coat, backfill then excavate the section after that. Just make sure when you backfill you compact the soil very well under the pipe to avoid any shear load between the repaired pipe and the original.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[BigInch]

Plan to be well on the safe side and support the pipelines at least every 15-20 ft. Since the work involves RE-COATING of those lines, severe corrosion might be automatically suspected and one might additionally presume that you may not have what we could call "entirely accurate information" on the pipe's remaining wall thicknesses to be able to safely assume that they will safely span maximum distances.

I hate Windowz 8!!!!

 

[Dario2002]

Thank You for valuable inputs. Corrosion on those sections is not deep as shown by last inline inspection (I'm not sure what is max depth right now, have no data on hand). Sections with higher corrosion will be cut out, and those with shallow corrosion will be only recoted to eliminate further growth.

As operating pressure, and design pressure are quite close (40/50 bar) I must check combined stress as recommended above. I will have to use some equation based solution. Any pointers for this situation?

 

[BigInch]

B31.4 or 8 both have combined stress equations. Will you have no pressure in the pipeline when conducting the work? Subtract the amount of wall lost to corrosion.

I hate Windowz 8!!!!

 

[Dario2002]

Finally I got back to this topic ..

As recommended I looked into ASME 31.8. Under section A842.2 "Strength considiration during operations" there is calculation for combined stress

However, I can not get grip on those terms:
[li] Mo out-plane bending moment, in.-lb (N·m)[/li]
[li] Mi in-plane bending moment, in.-lb (N·m)[/li]

In calculation those two are only used for resulting bending stress (Sb), which would be what in this case?
Can this resulting stress for straight pipe be calculated in usual way Sb=M/Z (moment divided by section modulus)?

If I take original formula from this standard: Sb=[(ii*Mi)²+(io*Mo)²]^1/2/z
and take intensification factors ii and io as 1, then I get that Sb=(Mi²+Mo²)^1/2/z which look similar to Sb=M/Z. Are Mo and Mi just transformed components of stress in cross section?


Just for reminder, model is straight pipe under pressure, which is dig-out and left on supports (earth or solid support).

 

[BigInch]

S_L is total axial (longitudinal) stress which is composed of tension or compression stress in the axial direction plus, or minus, the bending stress S_B.

Total axial stress =
S_L = P/A +/- S_B*c/I
P = axial load
A = cross-sectional area of steel pipe wall
M = I/c = Section modulus
don't forget any thermal loads, etc. when calculating P
making for
S_L = P/A +/- S_B/M



you must get smarter than the software you're using.

 

[Dario2002]

Ok, maybe I clear it up a little bit after Your input.

So I have this model:

For hoop stress:
sHP = P*D/(2*t)

For axial stress I need to take three load components:
sLB .. bending stress (M/Z)
sLP .. stress from internal pressure (SLP=Fa/A, Fa=P*Ai, A=pi*(D^2-d^2)/4)
sLR .. stress from restraint ends

This sLR I found to be = v*sHP-E*alpha*(T2-T1) .. (v is poisson's ratio, alpha is thermal expansion rate)
Is this correct?

Now following your input above, total axial stress would be:
sL = sLP+sLR+-sLB


And now I can use this in ASME calculation for combined stress?

sT (torsionall stress) is zero?
That would leave that combined stress is:
sC = sL - sH

??

 

[Dario2002]

actually last equation should return absoulte value: sC = ABS(sL - sH)

 

[BigInch]

SLP you say is pressure stress, but your equation is based on pressure on end caps, or what I call the "pressure on the end caps stress", so if you have full anchoring in the axial direction, that force is (theoretically) applied on end caps located actually outside the anchors, so stress is removed by the anchors before reaching the pipe between anchors, so it never makes it into pipe between anchors.

The thermal stress is an axial load in all pipes between anchors, but v (Poisson's ratio) is not for thermal stress. For thermal stress you use only [&alpha;], the thermal expansion coefficient.

The axial component of pressure stress is hoop stress (only) x Poisson's ratio.
SL_a = P*D/2/wt * [&nu;]

S_b is + or -

Don't forget to check the allowable for axial stress alone.
Additionally, since it is a thermal compressive load, check for column buckling.

you must get smarter than the software you're using.

 

[BigInch]

Forgot to say,
The axial component of hoop stress is similar to thermal stress, as it too only exists (fully) between full anchors.

you must get smarter than the software you're using.

 

[Greg80]

For such case, I would model it in Caesar.
However if you need a rule of thumb use table 121.5 in the B31.1

 

[uaepiping]

@BigInch,
"you must get smarter than the software you're using. "
Nice saying!!!

 

[chicopee]

A good reference on the subject introduce in the OP is "Theory and Design-Tubular Steel Structure", 2nd Edition, by James F. Lincoln Arc Welding Foundation, Chapter 9 Above Ground Pipelines. The second edition is my copy, however, updated editions may have been published.