Dear all,

At this time we are upgrading some of our seal systems from single to dual seal, mainly because of security improvements.
In applications of Hydrofluoric acid (HF) we have received proposals that specify a dual Type B seals (metal bellow)with plan API 53B.
Our experience in this applications is seal Type A with API plan 32/53B.
Some of you have experience with metal bellow seals in HF?
Advantages and disadvantages?
Can we to use only the API 53B without the API 32 plan?

You need the Plan 32 to prevent fluoride build up on the inboard seal.  If you are using carbon materials the flush will protect it from corrosion.  Plan 32 is normally, propane, alkylate or Iso Butane.  Make sure that the Alkylate is clean without HF contamination. When pump is on standby ensure flush is left on because you may experience pump siezure particularly at the back of the impellor where you have close clearances.  I have never seen a metal bellows in a HF application but thats not to say it will not work.  I suggest that you do a HAZOP on site to help determine if this seal is suitable.  My preference is always a  pusher seal in this application due to the modes of failure. The advantage of a bellows is that it removes the dynamic Orings but in the event of a seal failure (Ruptured bellows) it is difficult to control the leak, you should consider thisin the Plan 53B design, wheras a Pusher seal is not as severe perhaps.  Ensure that you design the seal with reverse pressure capabilities.  trust this helps!
What is the proposed barrier fluid?

Flexibox,
Thanks for your comments.
The proposed barrier fluid for the plan API 53B is isobutane, the same that we use for some other pumps with HF.
 

I've come across many metal bellows seals in HF service and have promptly changed them over to pusher seals when possible. You are very limited on the reverse pressure capability with the metal bellows, which is one of the reasons the pusher can be preferred in dual pressurized applications. The plan 32 is essential for isolation of the seal components from the acid. The plant I am currently at uses dual pressurized seals with a Plan 32 & 54 flush system. The Plan 32 fluid is isobutane; Plan 54 fluid is Alkylate supplied in a pump around loop by two dedicated Alkylate pumps. This configuration is a design specific to the unit licensor.   

I know of several plants that have gotten away from seal reservoirs and used a triple seal- a wet inboard and a pair of pressurized dry gas seals as containment.  All pushers.

I noticed that nobody mentioned using a Plan 52 unpressurized seal.  Good...  Iron fluorides from the flare header tend to end up in the reservoir contaminating the buffer fluid and causing seal failures.

We have a mixed bag in our HF unit: Dual pressurized pushers with Plan 32 / 54 in our pumps with HF streams and wet / dry 'triple' pushers in our pumps with trace streams - both run well. Yes, Plan 52 is a big no-no in an Alky unit IMO.  

bustox

HF service requires either killed carbon steel or Monel metal to combat against corrosion. Edge welded metal bellows from any reputable seal manufacturer i.e. AES, Crane, Burgmann etc will not be available. Whilst you may use a plan 32, you risk assessment should cater for loosing the flush.

The most common selection is a Monel pusher either face to back or back to back reverse pressurised and a Plan 53B. Very reliable solution.

Even with a pressurized dual, a plan 32 is typically a requirement to keep acid away from the inboard seal components. The face lubrication is handled by the barrier fluid; the 32 is there as another layer to isolate the seal from the HF. Most specs require it, but if you remove it that's on you I suppose.

Not sure where the above information about bellows was obtained from, but you can get edge welded metal bellows from most reputable seal manufacturers that will withstand HF service. However, they are not usually standard items and not typically recommended for reasons already stated; pusher seals are preferred.