WALL THICKNESS BASED ON DESIGN CONDITIONS OR HYDRO TEST CONDITIONS 2

[robs123]

I'd like to determine pipe wall thickness, may I know I should consider design conditions like design pressure, design temperature only or should I take hydro test (1.5 times design pressure) pressure in order to determine the pipe wall thickness?

This is to be used in skid fabrication were hydro testing will be done for 24 hrs @ 1.5 times design condition.

On which criteria we have select suitable wall thickness?

 

[zdas04]

Both. You probably have a stress limit as a function of SMYS. Generally a company stress limit is pretty low (I often see 20%) and that limit is usually applied to normal operating pressure (but sometimes it is design pressure). You need to come up with a wall thickness to satisfy that company policy. Then check that wall thickness against the code to make sure that your test satisfies the code requirements (e.g., in B31.8 you can't do a pneumatic test with methane as your test media above 72% of SMYS during the test). I've never had a pipe wall thickness that satisfied the company guidelines that failed the code limits, but I always check. When I used to do the calcs in the other direction I often found pipe that met the code but failed the company guidelines.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. —Galileo Galilei, Italian Physicist

 

[craghav]

I am assuming your design comes under ASME B 31.3 (Process piping in refinery), since there are certain allowances like corrosion allowance, usually pipes will sustain hydro load when they are designed for design pressure at design temp.
Use formula provided in 304.1.2 and Table A-1
Code also says piping should be designed for bending stress and other dynamic loads apart from internal pressure
Kindly go through the ASME B 31.3
After designing as per code, now check the pressure of Hydrotest and compute values and verify, if your design can not able to withhold hydro pressure, design for hydro pressure as well

Raghav

 

[robs123]

Thanks for your replies Sirs,

Yes I am designing as per ASME B31.3 process piping (formula provided in 304.1.2 and Table A-1) From my previous experiences many suggested me WT is determined only using design parameters like design pressure & temperature & not hydro test pressure. So now i am totally confused.

My inputs are material: API 5L GR.X42 pipe, design pressure 92 bars, temp 80 deg C, Line size: 48 inches, used to make Meter Skid in refinery. Corrosion allowance taken is 3mm. As per ASME B31.3 taking design pressure WT obtained is 42.6mm, By taking hydro test pressure WT obtained is 58mm, so which thickness is correct? & to perform hydro testing for 24 hrs for whole skid.

Please suggest.

 

[zdas04]

I think you are in the wrong code. For 48 inch pipe Grade X is 0.5 inch (12.7 mm) you need more than 4 times that wall thickness. You really should be looking at ship-rolled forms and BPVC calcs.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. —Galileo Galilei, Italian Physicist

 

[LittleInch]

robs123,

The esteemed gentlemen above may have confused the issue a little as you did not give all the information in the first post (design code is very important in these discussions).

To determine wall thickness to B31.3 you use design pressure as this is what it says in the code. The code already allows for the test pressure without breaking the pipe therefore although 24 hours seems a little excessive.

They are correct to point out that wall thickness to resist hoop stress is the first step and you need to check that the combination of pressure, temperature expansion /contraction stress and any bending stress do not exceed the allowable combined stress, especially when you add supports into the mix. Hydrotest loads are also calculated but use a different maximum allowable stress value to the normal operating mode. This is all normally done by the stress engineer. In the majority of occurrences the wall thickness to resist hoop stress from internal pressure is the one you end up with plus any corrosion allowances.

The worst case is normally gas lines where the pipe is designed to normal operating with gas, but then when you fill it with water and pressurise, it can fail / exceed the allowable stress for hydrotest.


My motto: Learn something new every day

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

 

[LittleInch]

I get 42.6mm wt as well to B 31.3. It's a big thick pipe fair enough, but not excessive.

My motto: Learn something new every day

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

 

[zdas04]

I just did a Google search and the thickest 48-inch pipe that I could find specifications for is 1-inch (25.4 mm). You are looking for something like 2-inch (50.8 mm). You have to get out of the piping codes for that.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. —Galileo Galilei, Italian Physicist

 

[LittleInch]

It's not going to be off the shelf, but can be made. d/t of 25 ish / 43mm is still piping code stuff - when you get to 70-80mm that's a bit different. Anyway that's the OPs problem, not mine - the key point is use the design pressure as your start point for working out what the wall thickness is in theory and then see if you can find someone to make it....

My motto: Learn something new every day

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

 

[zdas04]

I just looked at API 5L and the current (44th) edition is the first one that specifies a maximum wall thickness (52 mm). The OP's calc of 58 mm for the test case would make it not API 5L pipe. Even 42.6 mm is getting kind of close. For a vessel length, I'd walk away from this foolishness and call it a vessel since getting a mill run of 43 mm wall thickness 48-inch pipe would be a touch expensive.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. —Galileo Galilei, Italian Physicist

 

[robs123]

So for 48 inch API 5L Gr. X42 pipe with 42.6mm WT will withstand hydro test pressure of 138 bars (1.5 x 92 bars)? is that right?

 

[robs123]

Also do we need to add mill tolerance of 12.5% on it?

 

[LittleInch]

robs,

first question - Yes it will

second - If you look at 5L you'll find that for welded pipe above 15mm wt, you go down to +/- 1.5mm for wall thickness tolerance. B31.3 appears to add manufacturing tolerance to the calculated thickness tmin (304.1.1), but either way it's not much at this sort of thickness.

As zdas04 says, you need to find someone to make this.

Can't work out why you're going for X42 instead of A106 B - you get no benefit for the extra SMYS - both have allowable stresses of 20,000psi and is probably a bit easier to get.

Latest edition of API 5L is now 45th, but appears to be nearly identical to the 44th version.

My motto: Learn something new every day

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

 

[robs123]

As client suggested to take for flanges: A694 F42, Fittings: A860 WPHY 42, So corresponding pipes shall be API 5L Gr. X42, I have approached some manufacturers who can make this, but thickness not yet finalized.

 

[LittleInch]

Clients are not always correct - seems to me you're buying more expensive, longer schedule items (F42, WPYH 42 and X42) for no mechanical benefit compared to more standard pipe and flange materials. However if they're happy to pay and wait....

My motto: Learn something new every day

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

 

[robs123]

Yes this are required for high pressure transmission lines, can we weld API 5L Gr. B pipes with WPHY 42 fittings?

 

[LittleInch]

Can't see why not - It may need a separate weld procedure - I'm no welding expert so you would need to see if it fell into the same "group", but I suspect not.

If you're designing to a pipeline code then you get some decent benefit from higher SMYS pipe / materials, but very little/none from the same material when you use B 31.3.

My motto: Learn something new every day

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