Liquid Pipeline Thermal Relief 1

[lsusiegel]

I am the Maintenance/Engineering Manager at a bulk chemical logistics terminal in south Louisiana that handles many different hazardous chemicals. In particular, I am having issues with thermal relief on long uninsulated pipelines (over 5000' in length) in Propylene Oxide, Dimethylamine Anhydrous and Methanol services.

Following pipeline transfers of these products to and from our customers, we block-in the lines awaiting the next transfer. In all cases, we use Consolidated 19,000 series (3/4" x 1") hard seated thermal relief valves on the lines that relieve back into the storage tanks/vessels as necessary to prevent pipeline overpressure. In summer, these thermal relief valves may operate every couple of minutes (sometimes more, sometimes less) to relieve the line pressure.

I don't believe this type of valve is designed to operate in this manner or at this frequency as they are basically operating as a pressure control valve. We constantly have issues with the reliefs leaking by thus causing inventory reconciliation issues. At times, a new valve will last a matter of months in these services and must be replaced because they "beat themselves to death".

Are there alternative valve designs or other hardware (expansion tanks?) that I should consider in this application? I am meeting with Consolidated later this week for their recommendations.

Thanks

 

[pdprao]

It appears the thermal relief valve is doing its intended function. Probably what needed is to review the materials of TRIM and improvment of properties for longer life. A pressurized overriding tank with suitable compressible gas such as Argon/Nitrogen may be a solution which needs a competent technical review for implementation. It is being an issue pertaining to safety element of system ,standard code accepted methods with improved properties of TRIM could be a better solution. Consolidated perhaps will a give a better solution with their expertise with MOC of the trim. Clean conditions at seating surface also need to be assessed for possible improper seating due to deposits on seating surface.
Best of Luck

 

[BigInch]

First, Let's try the Simple Solution

Why not allow the pipeline's pressure do drop a little lower before you shut in to await the next delivery event. If you can do that at a low enough shut down pressure, so that the pipe's pressure remains below the valve's set pressure during a thermal pressure rise event, this problem would be easily solved.

Second,

It might be that the relief valve is just set too low, so first review the set pressures in relation to the maximum allowed pressures (MAP) of each pipeline. Relief valves can often be set higher than MAP in some circumstances to accommodate transient pressures, which may permit them to occur without activating a relief valve since, as long as such pressure rises remain less than the transient pressure allowable, are of short duration and not frequent in number.
It sounds like the problem is frequent enough that they may not qualify as transient pressures.

Lastly,
You may otherwise be forced to reduce the pipeline's operating pressure so that any rise in pressure still remains below MAP.

you must get smarter than the software you're using.

 

[LittleInch]

I'm taking a guess here that these are mainly Above ground lines? In that case it only takes a few degrees to raise pressure by a lot so the suggestion by BI may not work as well as if it was a mainly buried line. You don't say what size or volume you're talking about and I'm not familiar with the thermal expansion of those fluids, but if you can first work out how much liquid you're talking about to keep the same pressure when you go from storage tank temperature to static temperature then you've got a lot more idea about what is happening and how big you would have to make any expansion vessel / bladder.

You can get constant pressure bladders which would try and maintain a fixed pressure, below your TRV set pressure and hence be essentially empty when operating but allow the full volume of expansion when static. As no fluid would leave the metered system your balance figures should not be affected.

For pressure rises like this which occur quite slowly and involve small volumes you might need a much smaller orifice on your TRV or perhaps a different type all together. I've heard many TRVs "whispering" in a fairly constant manner on manifolds during hot days so your valves may just be opening too fast and letting go too much fluid in one go.

A very small control valve may do the trick also but would probably need a fine filter in front to avoid blockages.

From a look at the 19000 series it looks like it is designed to "pop" and relieve a set volume before re-seating rather than "simmer".

Please give us some feedback from your discussions so we get some learning as well.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[lsusiegel]

Thanks for the replies - all good information.

Yes, the pipelines are above ground. The PO and DMA pipelines are 4" schedule 40 carbon steel (150# class flanges) and the Methanol pipeline is 16" standard wall (150# class flanges). The PO pipeline volume is about 9700 gallons and the DMA pipeline is approximately 3600 gallons.

I will let everyone know the outcome of the meeting with Consolidated.

 

[BigInch]

Expansion bladders for such short lines? I doubt he'll get that budget approved.


you must get smarter than the software you're using.

 

[LittleInch]

He asked for alternatives, not cheap alternatives...

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[lsusiegel]

My experience with expansion tanks/bladder tanks is not good. In fact, I have one in my shop right now with a split bladder. But, that is indeed an option being considered.

Thanks - Mike

 

[StoneCold]

Based on the thermal expansion volume he might be able to add a vertical pipe with a nitrogen pad on it downstream of the flow meters if possible. Just set the nitrogen pad pressure regulator at just above the flowing fluid pressure. Then when you stop the riser should be empty, It will fill as the pressure begins to rise. He should be able to easily work out an acceptable vertical pipe volume (gas volume) so that the relief does not go off. Though obvious, I think it is necessary to mention he will want a quality check valve right as the nitrogen line goes into the riser.

Thanks
StoneCold

 

[BigInch]

Silly me. I always think simpler and cheaper, or no-cost, trump complicated and expensive.

you must get smarter than the software you're using.