Nexen Pipeline Failure

One reference says 100C, the other says 150C. Anybody that's done pipeline stress would find it difficult to believe that a buried pipeline could pass allowable stress numbers at that (those) temperature(s), yet the pipeline appears to be buried in the pictures they showed. In fact, at those temperatures (even at 100C) to get the pipe stress to pass, I'd think quite a few very general and conveniently, convenient assumptions would have to be made concerning soil properties, friction factors and especially their general lack of homogenuity. Also interesting to note that the failure apparently occurred after shutdown and restart, a period of extreme stress reversals for pipe-in-pipe (PIP) fabrications, what should be an especially worrisome time when the SCADA leak detection system should have been working, or if not, observers stationed along the pipeline.

PIP is far too labor intensive and costly for my tastes anyway.

The way these lines are constructed, they pre-expand the inner pipe by using the inbuilt heating system to heat the pipe up to about half way between ambient and maximum operating temp. The inner pipe expands and grows out the end then they weld it to the outer pipe via series of "bulkheads" to lock in the expansion stress. Start up and shutdown shouldn't affect it as the heating system should kick in to keep the inner pipe hot. Hence 150C is probably design and 100 the operating temp.

The thing I've never liked about this PIP system is the vacuum you need to pull on the annulus to get the insulation to work effectively. Monitoring this annulus pressure then becomes your leak detection system either from the inner pipe or the outer pipe. however any exterior or interior leak destroys the vacuum and the insulation properties.


Its not clear to me though what happens if you get a small or big leak in the inner pipe which then heats up the outer pipe in a small area. The outer pipe is not normally a pressure containing pipe, but as thin as you can make it to reduce costs. Thus it could easily fail under thermal stress.

A "fish mouth" failure implies local buckling and overstress to me.

Given the contents are a mixture of oil, water /steam and sand, the potential for erosion or even under sand corrosion seem quite high.

Will be good to read the report when it finally comes out in about 18 months time.




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

I can't find any reference to a vacuum in the annulus (as is sometimes used for liquid cryogenics), documentation seems to refer to temperature as the signal of a leak in the inner pipe.

Piping Design Central

I agree with the Little/Big Inch comments above, especially that a "fish mouth" failure implies local buckling and overstress.

The inevitable thermal cycling of the entire system could have broken something within..... the support scheme might not be functioning as designed...... There is no way to tell.

I also wonder about the design pressure/temperature rating for the exterior pipe. Shouldn't it be the same as the interior pipe ? What was the actual pressure test of the exterior pipe ?

The claim in the above link: "....provides a mechanically sound support for anchoring the inner pipe and thus allowing the installation of the high-temperature pipeline in non-competent soils with no expansion loops and no external anchors" bothers me...

I sense the smell of an MBA ...... with a strong concentration in sales and marketing

MJCronin
Sr. Process Engineer

GGator, I've looked at this system in depth quite recently and it was the one thing that surprised me but they need the vacuum to get the increase in insulation without needing an outer pipe which is much bigger than the inner pipe. Its not something you find in too much documentation but otherwise the insulation would be much thicker. It could be why the inner pipe might be 16" but the outer 20".

Remember this is a pipe with contents of 100c plus which would cool down rapidly if you didn't have very good insulation.

They do often run FO cable inside to check on temperature, but you might not see a big difference and its really there

MJC The outer pipe is often not a full pressure design to reduce costs as any internal leak is not supposed to gather. If the insulation gets wet its ruined so you can't hydro test the outer pipe and the inner pipe could collapse.

Mechanically it does the john normally but does impose stresses on the outer pipe to purees stress the inner pipe which might not have been assessed as well as it might.... Who knows, perhaps the outer pipe failed some time ago. I will be very interested to see the results of the investigation if they ever see the light of day.

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

I got some inside info on this a couple of days ago. Although no reports have been released, the info is that the temperature sensors were reporting high temperature in one section for over two weeks, but the operators did nothing about it. Given that the outer pipe is not supposed to see anything like this in terms of generate or pressure, it would seem that it actually tried to expand way beyond what it was designed for and has failed under a mixture of compressive stress and greater internal pressure than design.

All this is not "official" but comes from a very good source.

If this is correct, it sounds like it will be "operator error" rather than design failure.

Any one seen anything released officially??

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

It seems that a conclusion concerning root cause as design or operator fault would depend on exactly why and how the hot temperatures were generated, although continued operation at high temperature could not leave operators completely faultless.