Leak testing of jacketed oil pipe

[jtherm]

Hi,

I am looking for the best, most practical method of leak testing of steam-jacketed heavy oil pipe that is already field erected. Not only to prove that there is a leak, but how to locate it.

Actually, I suspect that there is leakage of the steam into the oil line that is feeding combustion burners at a weld somewhere in the piping where the steam jacket meets the oil pipe, which is causing some problems with combustion burner stability. The burner has a pulsing flame pattern when the steam is on, but the problem goes away when the steam is turned off.

There is about 200 m of piping between the tie in from the main, to the downstream process equipment, with probaby 15 to 20 welds where the steam jacket meets the oil pipe. Any suggestions greatly appreciated.

Thanks,
Joe

 

[BigInch]

I'd say you already have proof that there is a leak.

How to locate it?

Is there an unexplained temperature drop along the jacket?
Is an infrared or ultrasonic investigation possible?


BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[jtherm]

There is no abnormal temperature drop along the piping, but I suppose there would not be since the steam is not being lost, but entering the oil stream.

The jacketed pipe is also already insulated, so this is going to be a real pain for the end user to locate. They did all of the field piping, but I have my reservations about the quality of their work. The job location is in China.

Are you suggesting IR for measuring pipe temperature? I doubt that ultrasonic is available at this factory. I do not want to ask the customer to dismantle or redo what they've already done, but I am afraid our combustion system would not work properly otherwise.

Thanks for the reply.

 

[BigInch]

I was thinking that, if the steam loss from the jacket into the oil carrier pipe is large enough, you would measure (feel, or IR photo) a temp drop along the jacket wall. If the insulation is constant thickness along the jacket, at thermal stability, you should still see a temp drop on the insulation's OD.... if enough steam is lost.

If you know the rate of steam input and both the steam's inlet pressure & temp and outlet pressure & temp from the jacket, you might be able to calculate the theoretical pressure drop of the steam flowing through the jacket and check it against the actual pressure drop to see if there is some deviation which might be explained by a steam leak. Possibly even estimating the quantity of steam lost, but That could get a little fuzzy if process variables are not very accurately known.

Can you vary the flow of steam and oil? Turn down the oil and increase the steam flow. Maybe somebody can hear the leak from the jacket to the carrier without an ultrasonic black box.

They might even make those black boxes somewhere nearby. Supposedly it is accurate, so its worth the question.


BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[gerhardl]

Below mainly commeted before your last input on location and local difficulties, but anyway:

Sizes? Layout? Steam feeding pressure? Shut-off, draining and emptying possibilities both for jacket and inner pipe?

The leakage is probably of some size, although limited. A very minimum leakage could probably (by some burners?) be absorbed without problems (? -I am no burner expert).

The steam pressure for jackets are normally low, range 0.5 bar, but depending on pipe material and steam pressure, there is a possibillity of increasing leakage over time (corrosion,cavitation).

I agree with BigInchs conclusion, thermic camera if giving working results is the best and cheapest solution.

Other more complicated 'mechanical' suggestions:
If the steam jacket is equipped with steam traps: check for traces of oil, and location.

Visual inside pipe inspection if practical possible (size, layout and cost?) by pipeline cameras used for drinking water and offshore/onshore oil and gas pipelines)

Block off oil feeding pipeline both ends, but with pressure relief outlet (smaller needle valves or ball valves mounted). Jacket, with steam traps blocked for water drainage, to be filled with water, pressurized to jacket allowed pressure. Inspect for water at lowpoints or total in feeding line.

Or perhaps easier: same as above but jacket filled with pressurized air, closed and depressurized oil-line, manometer at oil pipeline to observe increase in pressure.

A more exotic and far-fetched solution: radioctive isotope incertion and tracing, perhaps at the oil side, to detect increase in outside (jacket) activity.

It could be difficult to measure steam loss, as uncertainity in data input would probably be greater than possible loss, and would need several measuring points. Theoretically possible, combined with thermic temperature readings and pressure readings along the jacket?

When you finally come to a conclusion, it would be interesting to know how the story finally ended and grade and type of leakage.

Good luck.

 

[lilliput1]

Does the burner use compressed air or steam for atomization?

 

[MikeHalloran]

Before you pressurize the jacket for leak testing, calculate the external pressure that will buckle the inner pipe.

... it makes a loud bang.



Mike Halloran
Pembroke Pines, FL, USA