FRP / GRE pipeline analysis

Hello everyone,

I need to analyze two GRE pipelines. The system is:
Underground with several changes in direction, with maximum straight length of pipeline about 400 meters.
Short radius elbows since vendor does not have higher radii bends!!!
Depth of installation is 2 m to top of pipe.
OD=157.8mm, WT=4.06mm, Tdesign=61degC, Fluid=water, Pipe qualified pressure=7.66MPa.
Soil is relatively good with 30 deg internal friction angle and 1800 kg/m3 density. Soil type close to sand and I model it using American Lifelines Alliance method sand/gravel type.
I use ISO 14692 for the analysis. I have populated all the cells with appropriate data supplied by vendor(see attached model and files).
I have also prepared several load cases(see attached).
The problem is that I have overstress and need to find a solution how to struggle with it.
I cannot use higher radii bends since they are not available.
To have code compliance I must shorten the maximum straight runs to less than 10meters!!!(I found it thru iterations). I cant afford this because vendor informed that we cant use more than 50 fittings per pipeline route!!!
Any advises would be appreciated!

Regards
Zhambul Igizbayev

Is the 61 C a real temperature? try to obtain to use max operating temperature instead design temperature.

Dan,

The operating temperature is 31 degC and even with it the system fails with 122% overstress.

With design temperature 61 DegC system fails with 140% overstress.

Client specification also restricts me saying: Piping flexibility and stress analysis shall consider the most severe specified temperature
and coincident design pressure.

Line list design temperature is 61 degC.

Additional Information: The pipes are with threaded connection.

Regards
Zhambul

Exactly:
Piping flexibility and stress analysis shall consider the most severe specified temperature and coincident design pressure.

Doesn't mean that you have to use design temperature.
Most probably your design is operating temperature + 30 C (31+30 C).

Discuss with client if is a real temperature or only a number in the line list.

Dan,

As I told the systems anyway fails with operating temperature 31 degC and the overstress is significant - 122%.

Line list was developed by our design company and we were guided by Client specification which stipulates adding 30 degC to operating temperature to obtain design temperature.

Regrads
Zhambul

Then maybe you can review the installation temperature or you place some anchor blocks.
You mention: Pipe qualified pressure=7.66MPa. 76 bars is for me very high for fiberglass.

Dan,

Thank you very much for your contribution. Much appreciated!

Gents please any other thoughts on the matter.

Who has good experience in design of GRE pipelines.

What about trust blocks at elbows? Who can advise how to implement them in the model?

Who dealt with threaded GRE pipelines? What are the weaknesses of such pipelines? What are pitfalls?

Regards
Zhambul

See the webinar:

https://www.houseofengineering.com/courses/wb-rec-dynaflow-08

https://www.dynaflow.com/training/webinars/

Thank you Bob.

I found interesting thing while was watching the webinar. Higher stresses occur at pipe adjacent to bends and not at the bends!
And results of my analysis prove that!

Regards
Zhambul

Dan,

The design pressure is 38 barg and the hydrotest pressure is 1.5*38=57 barg. That is why we have been suggested using such rating for pipe by vendor.

Regards
Zhambul

Can anyone advise or may be have experience how to model trust blocks at GRE elbows.

The difficulty is that it should be monolithic block of concrete and where the restraints should be applied to model it is not clear.

When I try to put anchors at the elbow itself and at some short distance away from elbow at pipe the software shows very huge overstress at this location between two anchors.
When I try to anchor at elbow ends I also have overstress but not huge.

The question is if it is an artificial overstress... thus need to find better ways how to implement trust block.


Please share your opinions.

Regards
Zhambul

Hello,

How about trying to use a thicker wall at the failure point to have a higher section modulus.

Warm Regards,

Borzki,

Thanks for contribution into the discussion.

Yes, I have already requested higher rating pipes data from vendor, now I am waiting for information to include into analysis.

I have also tried to increase WT with no change in allowable stresses and it helps.

If the WT increase does not improve the situation significantly wanna try increasing the OD of pipe in high stress locations.

Regards
Zhambul

Hi Zhambul,

No problem. That's essentially what we have done in one of our pipeline project where we requested higher wall thickness at those high stress concentration points that exceed the code allowable stress. Cost wise will not be so high impact since you do it locally only not globally.

Warm Regards,

Hi Borzki,

I appreciate that since you share a real experience.

Most likely I will do the same i.e. increase WT.

Thanks a lot.

Regards
Zhambul