Dear all,

For thermal bowing study of flare header from PSV to knockout drum. I have some questions as follow:

0. What is purpose of flare piping?
1. What should we consider for support study of flare header (support type and layout for big pipe size)?
2. What data should be get for thermal bowing stress analysis (different temp.)?
3. Do we need to consider slope of big size in the model (1/500)?
4. How will we input data in the model to simulate this thermal bowing effect?

I hope someone can help

Best Regards,

0. What is purpose of flare piping?
All Refineries and some Chemical plants have highly flammable gasses that may be released through a PSV (Pressure Safety Valve) when the PSV set pressure is exceeded by the system pressure. All of these hydrocarbon and Chemical PSVs are (or should be) piped to a collection header which is routed to a Flare Knock-Out Drum and then to a tall Flare Stack where the Gasses are safely burned off. The Flare Knock-Out drum provides a place for any liquid condensables to be removed and pumped back to the plant for reprocessing. The Flare Knock-Out Drum is normally located about 300 feet (100 meters) from the Flare Stack. A Flare Stack is capable of generating very high and intense heat for long periods of time during a full plant shut-down. So it should be located as far as possible from all other plant equipment and facilities as well as any public homes, shops, schools, hospitals, etc.

1. What should we consider for support study of flare header (support type and layout for big pipe size)?
The Flare Header is normally a large line, 24" (DN600) or 30" (DN750) or larger. This line would normally be supported above the pipe rack on the centerline of one of the pipe rack columns. The pipe rack column is extended up to form a "T" pipe support at a graduated elevation so the line slopes toward the Flare Knock-Out Drum. The top bar of the "T" must be very wide to allow for line movement and have tall and very strong "Stops" to prevent the pipe from jumping off the support.


2. What data should be get for thermal bowing stress analysis (different temp.)?
Do an overall Flare Header Isometric Diagram and locate all the hot discharges. Then do a temperature degradation study on the hot discharges. Don't get too finite with it you will not be able to determine if the release is during the heat of a mid summer day or the dead of the coldest winter night. Keep it a general condition study.

3. Do we need to consider slope of big size in the model (1/500)?
Yes you need to slope the Flare Header from in the plant (farthest & highest) point down to the top of the Flare Knock-Out Drum. Slope the Flare Header from the Drum up (at the same 1/500 slope) to the Flare Stack.

Good grief.

Have you never heard of employing someone with knowledge and experience ??

Jop, your comments are patient, well made and helpful ..... but regarding the questions put, a little knowledge is a dangerous thing. The guy needs in-house help.