Hello

There has been a failure on a 4" GRE hubless (heavy duty)flange to one of our offshore Drill water systems. Pipe and fitting material is Bonstrand series 7000M.

I am looking for the following information:

- Allowable load for a GRE flange joint (don't tell me to ask for the vendor)

The temperaure limit for the GRE 7000M pipe is rated to 93 degree C, but the flow temperature was reported to 115 degree C when the failure occured, even though line operating temp is ambient. I suspect that higher temp may have caused a failure on the joint.

Your advise and any information would be most appreciated.


Regards
STC

Hmmmn. How did it fail? What were the symptoms and the environment? High vibration? Extra-long service maybe? Any chance it (the joint) might have been abused (by accident) or an "undeclared" high side load from shock, impact from a crane load or nearby construction?

http://pipespecialties.com/pdf/FP196A.pdf

That's a writeup for Cl 7000 flanges - nothing really stands out as unusual or different about the installation that would cause a problem: torque limts and stud assembly requirements are not out-of-the-ordinary, for example.

One other story on-line about Bondstrand 7000M says its good for 121 deg C max operating temperature, so your 115 deg "violation" of a 93 deg C spec isn't even out-of-range for the material.

http://www.bondstrand-marine.com/_DOWNLOADS/PDF/FP918.pdf

The second source has technical spec's and material properties for the "pipe" and max span limits based on 21 deg C properties - but it didn't say anything (I saw at least) about maximum "side loads" allowed on the flanges joining two random sections of the pipe.

Thanks for your input and the relevant information.

The existing CS suction/discharge (6x4) line of the centrifugal vertical in-line pump was replaced by GRE pipe around two years ago. The pump nozzle flanges in CS joins the GRE flanges where the failure occurred. The suction pipe work (GRE) has come apart forcing the pump to move on the base plate towards the discharge. The pump only rests on its base plate without being secured to the base plate as per the vendor requirement. Rather, the suction and discharge piping (GRE) have been supported instead.

We don’t have much information at this stage except that the value of vibration was reported to be10 mm/s on the pump. It is my opinion that high temps may be the main cause for the failure, and high cycle fatigue from the pump vibration a possible secondary contribution.

Regards
STC

10 mm/sec vibration? If I'm calibrated right, that's be just under 7/16" movement? They shouldn't have reported that big of a movement, even if the pump weren't restrained. (Frequency would be cycles/sec; movement in mm (or preferably, in mm x 10^-2 or so.) Maybe your source isn't passing along good data ??)

I'm not convinced any flange on any pipe could hold a moving pump (big enough for a 6"x4" suction) if the pump weren't mounted firmly and accurately to grade. Particularly if he's describing mm's of movement.

Some firms design practices for small sized vertical inline pumps allow them to be supported in the connected piping or simply rest on a support.

This is ok when the connected piping is METALLIC and has sufficient strength.

It is not surprising that the GRE piping this dicsussion has failed if the existing steel piping in this vertical in-line pump configuration has been replaced by GRE piping.

I prefer to run steel piping on either side upto the valves in steel and then transition to a GRE material.

As a remedy work, we are going to secure the pump by bracing the pump flange to the base plate. If this does not work, we are considering the GRE pipe to be replaced to the metallic upto the valves as you advised.

Thanks for your help.