Metal Bellows in Flashing Hydrocarbon Services 3

[HoustonMechEngr]

We have metal bellows installed on plant-wide flashinig hydrocarbon problems and are having issues with our plan 52 pots pressuring up during operation. I've been told "spring loaded" pusher seals are a better choice for flashing hydrocarbon services. Can someone please summarize the overall advantages of the pusher seal vs. the bellows design?

 

[flexibox]

Metal bellows seals became very popular in the late 1980's early 1990's and were being specified on most applications. The reason was because they had no dynamic Orings. Many site experienced problems with bellows failing "factured end rings" on light hydrocarbon services. Then there was a shift not to use metal bellows in these services. See latest API 682 clause:

Metal bellows seals in flashing service are prone to fatigue failure, induced by “stick-slip” if marginal vapour
suppression occurs. If metal bellows are desired, the seal should be a totally engineered sealing system with
special attention to vapour suppression under all operating conditions of the pump, such as start-up, shutdown
and plant upsets.

If you are not experiencing fatigue failures with your bellows then I would check the balance ratio of the seal and spring loads. With a pusher seal the balance ratio/spring load may be higher than your current seal design. With a higher balance ratio and spring load you will get a tighter seal but may sacrifice seal life due to wear. This is a generalized comment not specific to your application.

It would be good if you can advise the Pumping temp/SG/VP/chamber pressure of the seals you experience this problem.

 

[bk19702]

We should clarify the balance ratio comment, as the balance ratio for the metal bellows seal will change depending on the process pressures being discussed.

Metal bellows seals are considered to be inherently balanced based on the forces that act on the bellows convolutions. The balance diameter of a metal bellows seal is the mean diameter of the metal bellows; however, at higher pressures the balance diameter will 'shift' due to the forces acting on the internal convolutions. So, at higher pressures the balance ratio would increase.

Balanced pusher seals in light ends service will have a balance ratio controlled based on the face geometry; in general the spring load will be lower on balanced seals vs. unbalanced seals, which enables them to be used at higher pressures and marginal lubrication regimes. In reality using the balanced seal may result in slightly higher leakage rates.

So while at 0 psig the balance ratio on the metal bellows seal may be lower than that of the balanced pusher seal, at 150 psig it will be a different story.

 

[gyaneshwarp]

Metal bellow seal or pusher type face to face seal with Plan 53 C should be ideal for this application.

In Plan 52 liquid does vapourise and sometime even takes the buffer fluid along with it.

 

[JJPellin]

We should thank flexibox for an excellent and well supported reply. I have a few comments to add. You describe a problem with seal pots building pressure but do not describe the failure mode of the seals. Are you using a true Plan 52 with the seal pot continuously vented to flare (or some other safe vapor control system)? If you are not venting the seal pots, then any seal will eventually cause pressure to build up in the pots. If the pots are vented and pressure still builds, then you have a considerable seal leak and should be replacing the seals. The failed seals should be exam mined to determine the cause. If the bellows are fracturing, then the slip-stick phenomenon is the likely culprit. However, if the seals failed from dry running, abrasive damage, plating on the seal faces, blistering on the carbon or grain pull-out, then the selection of bellows versus pusher may not be the key decision. If the service is dirty and hang-up of dynamic elastomers is a possibility, a pusher seal could make the problem worse.

A bellows seal, properly designed and installed can run very well in flashing hydrocarbon service. I installed an early generation dry-running containment seal with a standard bellow seal as the primary in 1991 in sour propane service. It ran continuously for nine years with monthly emissions monitoring. It can be done. But, if the same service came up today, we would choose a pusher seal because that is the preferred technology for the service. We would have a better chance of consistent long-term reliability with the pusher seal.


Johnny Pellin