Stress Analysis Design temperature for Multiphase fluid 4

[jonjose123]

Hi All,

We have a 8" pipeline material API 5L X65 and 5 km buried line.

Design temperature is 105 deg c where as operating temperature is 75 deg c.

We have question on what shall be the temperature to be used in analysis. We have used 105 deg c and the system shows over stressed on bends. Is it possible to use 75 deg c instead of 105 in analysis?

Buried soil temperature at 1 m cover depth is 28 deg c.

Expert advise on this matter is awaited from members.

Regards
B

 

[LittleInch]

In general I try it with design temperature and if it doesn't pass then start to look at other options.

One is to make sure your installed temperature is correct for you location - often a low conservative figure is chosen, but 28C seems to be good for you.

After that then yes, use max operating temperature. B31.4 31.8 and ISO 13623 all specify max metal temp during operation as the basic stress analysis for restrained lines.

Of course you need to then ensure that your max operating temp doesn't creep up and hence set suitable alarms on your temperature and suitable notes stating maximum operating temperature due to pipeline stress issues.

LI

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[jonjose123]

Thanks LI for your reply.

It was helpful.

As here the operating temp is 75 ged c. We have run the system with operating plus 15 deg and in that case the system is passed. However, if we raise the temperature to 105 deg c the system fails at bends.

Is there any option to reduce the stress at bends. Is it possible to increase the cover death at bends or use different backfill at bend area to reduce stresses... will it work..

Regards
B

 

[XL83NL]

Technically, I'd say the temp. for stress analysis is 105 deg C. Because that's what your design temp is. Although it will not be serviced continuously at that temperature, and perhaps may never see that temperature, the pipeline needs to be designed as such due to the design condition. But then again, Im a B31.3 engineer so not 100% sure if LI's method is a general accepted method under pipeline codes.

I therefore, for B31.3 piping, use the PID design temperature (as shown on spec breaks) for stress analysis. Though conservative, it's more difficult to prove or justify the opposite, i.e. whatever reason you have to use a lower temperature for your analysis. In the end, your PID says your system should be designed or rated for that condition. That method fits imo best with B31.3 para 301.3. As youre using pipeline, there's no design temperature definition under B31.4/.8, but you should be save with this method.

If you deviate for whatever reason, just document why and how you verified that a lower temperature is safe to use for analysis, in the case that for a future event someone needs to review this.

 

[pennpiper]

In Real Estate there it's, Location! Location! Location!
In Piping it's Temperature! Temperature! Temperature!
The key is which Temperature!
Re: Peng's, Pipe Stress Engineering 1st Edition, 2009.
On Page 29 there are definitions for Temperature.
- Operating Temperature - "This is the normal operating temperature"
- Design Temperature - "This is the maximum sustained Temperature during operations"
- Flexibility Temperature - "The temperature to be used in thermal flexible analysis"
-
The Design Temperature listed above is not the "Hypothetical Design Temperature"(HDT) used by Process Engineers and Piping Material Engineers to establish the pipe wall thickness.
Peng does not identify or discuss this HDT. This HDT is often one and a half times (or more) the maximum sustained Temperature during operations and therefore grossly exaggerates the results of the calculations of the analysis. It is never seen by the commodity or the piping so why base the piping configuration and cost on a hypothetical?

Just my opinion:
"Stress analysis should be based on real (expected) temperatures not a hypothetical (never experienced) temperatures".

Sometimes its possible to do all the right things and still get bad results

 

[LittleInch]

XL83L Therein lies some of the difference between piping and pipeline design.

Process Engineers have an annoying tendency to add certain margins to max Op temp ( often 20C or more) and although it causes a few issues within piping stress design, the material is thick, relatively unstressed and can cope with it.

Pipelines are much more sensitive and hence that's why the design codes only talk about max operating temperature for thermal stress ( I checked before I wrote that...).

Often the only reason they have such high design temps is that they connect o piping which has an inflated design temp and it's difficult to make a drop without putting in place trips, coolers etc t deal with a situation which doesn't exist.

However if possible the max design temp for the pipeline should be lowered to no more than 5 C above max operating and then everyone is happy, now and in the future because then it is fixed in the design basis / P & ID etc so no one runs it at 100C in the future.

Options for overstress at bends include:
Thicker bend and bend tangents ( 1-2m after the bend )
Shallower radius (10D or 40D if you've got the space
Maybe split a 90 into s 45's separated by a length of pipe (can sometimes make it worse....)
Compressible backfill can work, but difficult in practice to build.
Cover depth normally doesn't do too much for you, but give it go.
I once had to allow for an open box at one bend, but that was >110C and in desert.

Anything more than 100C and you'll get a steam geyser when it rains....




Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[jonjose123]

Thank you XL83NL, pennpiper and LI for your valuable suggestions.

As suggested by LI , I will try run the system maintaining the temperature at 105 deg c.

Also as suggested by Pennpiper, Will check with Process Guys to provide the real expected temperature for flexibility analysis.

Hope it goes well.


Regards
B

 

[pennpiper]

Go here for related discussion:
pipingdesigners.com/forum?view=topic&catid=54&id=1499

Sometimes its possible to do all the right things and still get bad results

 

[KevinNZ]

In our piping designs to B31.1 & B31.3 it is very clear we design to "design temperature" often the design temperature is related to over pressure protection or other safety setting. The deign temperature is calculated knowing what is going to be used for.

In piping design it just as important to know the min design temperature. The expansion stress range is calculated from min to max temperature.

For your pipeline you could try
[ul]
[li]Thicker wall elbows[/li]
[li]Refine the soil strength factors used in your design[/li]
[li]Shorter runs between your elbows.[/li]
[li]Anchor blocks at shorter spacing to reduce the expansion at the elbows. Will put pipe runs into compression. But I am not a pipeline designer and this may make more problems[/li]
[/ul]

 

[LittleInch]

KevinNZ

I appreciate piping is different to pipeline but then so are its conditions.

For an upset condition which Design temp is there to accommodate, you could easily get metal temperature in say a 50m section of pipe that you won't get in a 500m of buried pipe.

Its also a lot harder to get a good and fixed installation temperature because it is so variable - installation below ground is very stable.

BTW in pipelines we call changes of direction "bends" not elbows.... Bends can be anything from 1 to 90 degree (I've never seen more than 90 but you could have 90+) and radius anything anyone can make. Commonly 3D, 5D, 10D and 40D.

You would need bends every 100 - 150m to make a difference and the cost in bends, fabrication and increased length of pipeline is horrendous.

Ditto, anchors just cause more problems than they try and solve - nearly always a bad move on pipelines.

Thicker bends and their tangents and more springy soil can make a difference.

Bottom line - try design temp and if it doesn't work get a max operating figure you can get confirmed and then use that. if that doesn't work then try some other things....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.