Low Pressure Steam Bellow Failure 3

[Designer80]

I have a question.
Recently one of our Low press Saturated steam ( 6.5 barg) mettalic expansion bellow got ruptured.when steam was isolated for few minutes (30 mins) and then again opened through crack opening the manual valve (shown in the sketch).A hammer sound appeared and bellow ( encircled in the sketch) got ruptured.We have checked any possibility of condy logging through reverse flow of condensate in the steam line which could have resulted hammering and consequently failing the bellow.
Can you please share your experience, what could have gone wrong and what precautions we can take in future?

 

[Designer80]

There are also slip-style expansion joints for steam that can be used instead of bellows. I personally LOATHE bellows expansion joints, for exactly this type of failure.

First choice - if there's enough space - expansion loops. Second choice, slip-style. If there's absolutely no other choice - bellows.

I would urge everybody to look at the articles & links on the Kirsner's website I linked above. I would also highly recommend "Hook-Ups - The Design Of Fluid Systems" by Spirax-Sarco. It was my go-to steam system book for decades, and it never let me down.

We have made some progress on the matter.
Historically, this FCV had a minor steam passing (1-2 te/h ), normal steam flow through this FCV is 85-90 Tes/h.At the time of incident, team had isolated manual gate valve (24 inches ) to stop steam flow towards reboiler ( because distillation column was supposed to be shutdown ) Condensate line is connected to a common condensate drum in boilers area which contained an NRV ( check valve) to prevent reverse flow. WE doubted that NRV as well but it found intact.When steam Manual valve was close there was still some cold process condensate in the column which was being rained through the column into bottom sump.This cold process fluid may have created Vacum on reboiler steam steam side ( due to condensation) which consequently could have lifted the condensate from the pot ( shown in the sketch) and filled the steam line till manual IV.Please correct if there are any technical flaws in this theory. Thanks

 

[pierreick]

Hi,
Attached a document to support your work.
BTW, is this the right type of manual valve?
Pierre

 

[Designer80]

Hi,
Attached a document to support your work.
BTW, is this the right type of manual valve?
Pierre

Thanks. Indeed it will be useful reference resource book

 

[LittleInch]

You could be right, but don't get tracked down into finding ways to excuse what happened. The key error here, IMHO, was using a gate valve to isolate between a live steam system and cold steam system. We don't know if there was condensate on the upstream side or downstream side of the valve, but it should never have been used to re start the system. It is simply far too brutal and uncontrollable and the steam passes through at the lowest possible part of the pipe as you crack it open.

In retrospect if the process could handle it a small bleed leakage was probably the best thing to do via the FCV to keep it all warm, pressurised and able to slowly ramp up under control. I can understand why someone thought it was best to isolate the steam supply, but I doubt this was ever envisaged as the operating philosophy or procedure on shutdown and restart of this system?

 

[Designer80]

You could be right, but don't get tracked down into finding ways to excuse what happened. The key error here, IMHO, was using a gate valve to isolate between a live steam system and cold steam system. We don't know if there was condensate on the upstream side or downstream side of the valve, but it should never have been used to re start the system. It is simply far too brutal and uncontrollable and the steam passes through at the lowest possible part of the pipe as you crack it open.

In retrospect if the process could handle it a small bleed leakage was probably the best thing to do via the FCV to keep it all warm, pressurised and able to slowly ramp up under control. I can understand why someone thought it was best to isolate the steam supply, but I doubt this was ever envisaged as the operating philosophy or procedure on shutdown and restart of this system?

Thank you. Actually same practice we had been doing since atleast two decades and nothing happened (two years back during same activity, manual gate vapve was kept closed for 3 hours before it was craked open whereas this time it was mere 40 mins) thats the reason we are trying to find what has changed this time which led to failure. Am 200% agreed with your analysis about culprit gate valve.

 

[LittleInch]

Well you didn't say that before now.... So maybe some extra cooling from the re-boiler had an impact, but either way, that procedure looks to me like you've been getting away with it rather than it being a good idea.

Maybe this time the "cracking" was bit more violent / faster than before?? That's the problem with something like that - controllability is very limited.

 

[TBP]

From the information provided, I don't see the gate (or any other style) valve in and of itself being the issue. I believe the root cause of this event as being the slug of water, from whatever source, in the steam line. If the steam trap ahead of the gate valve is correctly located close to the main line valve - and it was functioning as properly - that should have taken care of any condensate building up in the main line upstream of the closed gate valve. That would mean, from the info & sketch provided, the water was in the downstream section of the steam line. The source of that water needs to be identified, and the issue corrected before this event repeats, possibly with more dire consequences.

I strongly suggest a free-blow drain line be installed just downstream of that big gate valve, and that the operating instructions modified to have that drain valve opened - and left open - whenever the main line valve is closed.

 

[LittleInch]

I kind of forgot that this was only closed for 30 to 40 minutes, but that seems to be enough to cause the system restart to revert into a full scale blow down situation when you need to go through the same procedure for a 30 min shutdown as you would for a 30 hour shutdown.

 

[Designer80]

From the information provided, I don't see the gate (or any other style) valve in and of itself being the issue. I believe the root cause of this event as being the slug of water, from whatever source, in the steam line. If the steam trap ahead of the gate valve is correctly located close to the main line valve - and it was functioning as properly - that should have taken care of any condensate building up in the main line upstream of the closed gate valve. That would mean, from the info & sketch provided, the water was in the downstream section of the steam line. The source of that water needs to be identified, and the issue corrected before this event repeats, possibly with more dire consequences.

I strongly suggest a free-blow drain line be installed just downstream of that big gate valve, and that the operating instructions modified to have that drain valve opened - and left open - whenever the main line valve is closed.

Thank you, Whats your opinion about the steam trap located just after the manual 24 inches Gate vale. Will it be sufficient to drain the condensate from the line or it will become non-functional if there is only condensate in the line and no steam (when manual Valve is isolated)?

 

[TBP]

While there are several different styles of steam traps, at the end of the day, they are just automatic valves. They open in the presence of air or water, and they close when there's steam. That's it. The amount of condensate (or air, on start-up) they pass is based on the size of the orifice, and the differential pressure across it. When starting-up a steam line, the first thing the trap has to eliminate is the atmospheric air in the line, which is more dense than steam, and will be pushed out of the bottom of the pipe, along with the large amount of condensate that forms from just heating up the pipe. This is all happening while there's typically very little - or no - differential pressure across the trap. Sometimes, because of system layout and elevation changes, there will be back-pressure on the condensate system that must be overcome before the trap can discharge anything at all. In these cases, there should be manually operated drains, that are left open during start-up. As the line heats-up, pressure will build to the point that the traps will work - the air has been discharged, and the pipe material has heated-up substantially, so the start-up (relatively large) level of condensate has been dealt with. At this point, the drains can be closed, and the traps should be able to carry the operating level of condensate, discharging it to where you normally send it - typically back to the boiler feed system, or in some cases, dumped to drain.

 

[RVAmeche]

Yeah IMO the biggest drawback of bellows expansion joints are:
1) People forget you can't just throw them in. Supports etc have to be reviewed
2) A lot of places forget to put them on a maintenance/replacement schedule or ignore it and just let them eat. Eventually they fail

 

[LittleInch]

That was my point in a roundabout way - what is the procedure for a start up from cold in this section?

And was it followed on this occasion or because it was seen only as a short term isolation - "isolated for few minutes (30 mins) " that the full start up procedure wasn't followed?

A "few" minutes to me is 5 or 6, not 30.