From napkins to fabrication VFT boiler design

[fredrosse]

Mike, "When we closed the valve, the valve failed through the stem going on through the moving part."

I know I am fairly thick at times, but I can't figure out what happened here? I thought the stem was the only moving part?

Fred

[S. Weaver]

@Fred: Can't speak for Mike but some bigger valves are fit with a rotating seat. These seats have been known to come loose or off or became dislodged somehow leaving a passage for steam.

@Greg: That's correct. In applications where the tubes are short, some have welded them in with full penetration. It's still not the best practice.

[Lopez Mike]

I just grabbed the broken valve cup off of the shelf and will take some pictures of it and post them in a day or two.

I tend to save things like this just to keep myself humble and careful. I once started a new job by tearing the top off of zip lock bag. I kept that bag top in the pencil drawer of my desk for a long time just for a morning reminder. "You too can be stupid!"

[87gn@tahoe]

The fatal railroad incident with the "unsafe" valve involved the victim trying to open a valve with a wrench on the handle that was already open (he thought it was stuck closed apparently...). In that design the lack of a separate collar allowed the entire assembly to blow free of the valve body, hitting the victim and scalding him to death with hot water and steam.

No matter what type of valve it is, NEVER EVER adjust anything on it while under pressure other than the packing gland.

[fredrosse]

Back to the 26 April posts, and BOILER STAYS, some additional information:

One statement”

“A stay tube is simply a tube that is attached to the tubesheets to support open areas of the tubesheet that are NOT supported by the normal firetubes. The boiler design you are developing (Stanley type VFT) would not require any stay tubes.”

Followed by another:

“In time the tubes will corrode and you wouldn't want to rely on them for any sort of boiler strength. A failed tube or tubes should result in a leak, not the tube sheet bulging"

In fact, thousands of boilers rely on the tubes themselves as the only piece resisting the bulging of tubesheets. Only when there are portions of the tubesheet area that is not supported by tubes are stays required. The ASME code has all the rules and criteria for this, however a typical Stanley type boiler would not have any stays. For a VFT with integral furnace, there are often stays between the outer shell and the furnace tube, and sometimes vertical stays supporting some portion of the top tubesheet that does not have the firetubes support.

The reason why this is acceptable is because the tubes will start leaking long before they jeopardize their tubesheet support duty. Allowable tube materials are always ductile material and will not be prone to cracking. (As a side observation, this is why the ASME Code prohibits stainless steel tubes for any power boiler)

As an example, examining the Margaret S. Boiler (16 in shell diameter, 48 tubes, 1-1/4 x 0.095 wall) at 250 PSI steam pressure. Assuming the tubes take all the pressure forces that tend to put the tubes in tension (no credit whatsoever for the tubesheet to outer shell welds), the tube stress in tension is only 1870 PSI, compared to the minimum yield stress for this material of 26,000 PSI. With uniform reduction in wall thickness, and 250 PSI steam pressure, the tubes would yield in tension at a thickness of 0.007 inches.

Of course, as tubes corrode they do not do so uniformly, small pits occur, and some areas of the boiler corrode more than others. It would be a cold day in the lower regions of this earth when all those 0.095 wall thickness tubes corroded down to a wall thickness of 0,007 (thickness of a few sheets of ordinary writing paper) before one (or several) sprung a leak. It is clear that the tubes would have to be replaced long before their failure in holding the tubesheets firm would be a concern.

So, in conclusion, tubes acting as stays are very effective, and will leak long before their strength at holding the tubesheets from bulging becomes a concern. Open areas of the tubesheets that do not receive support from the tube bundle may require stays, but generally not on a Stanley type VFT boiler.

The attached picture shows some acceptable tube to tubesheet attachment methods.

[johngriffiths]

Hi Folks,

At least one other UK based forum member will know of an incident on a passenger carrying vessel where the bonnet of a valve under steam pressure unscrewed and blew off. The valve was remote from the operating position and the operator could not see what was happening. The owners have since changed the valve type.

An ordinary bonnet can be prevented from unscrewing with a suitably designed locking plate but it is better to change the valve.

Cheers

John

[gondolier88]

Hi Folks,

At least one other UK based forum member will know of an incident on a passenger carrying vessel where the bonnet of a valve under steam pressure unscrewed and blew off.

John

Indeed John, there is...!

The culprit; a PN16 rated flanged valve on SY Gondola that had been installed for at least 10 years and had given no trouble at all, it's service was boiler pressure steam to the bilge ejector.

All crew on Gondola are trained to operate the engineroom safely, with all of us having varying degrees of steam experience.

On the day of the accident the crewmember in the engineroom was the person with the least steam experience- although this was not the reason for the accident, I'm sure that it contributed in effect; the valves had been lagged in their entirety with lagging 'bags', with only the spindle protruding. Upon rotating the spindle to open the valve, or so he thought, the bonnet was actually the moving part- with 120psi on the gauge the bonnet came loose- pushing the crewman's hand against the deckhead with a painful bump and filling the engineroom with enough steam to completely impair vision in around 3 seconds!

'Luckily' the valve was on an isolateable manifold- this meant there was no priming.
'Unluckily' the valve was on an isolateable manifold- but the isolating valve was only accessable by reaching obver the valve that had failed- a more experienced member of the crew who was not suffering from shock put on two waterproof jackets, one over his face and managed to access the isolating valve enough to be able to safely limp back to the jetty.

Lessons learnt;

- As a rule, try to make all valves in service of a minimum PN32 rating, preferably PN40- both of which under BS/EU standards now have union bonnet connections, meaning that if they came loose they cannot blow off, they would leak enough steam to tell you there was a problem- the one installed now is PN32.
- Lagging is great- but not to the detriment of accesability or safety.
- Crew training is essential- the crewman involved immediately exited the engineroom and didn't try to be a hero- other crew onboard knew how to tackle the problem and how to administer first aid to the crewman involved.
- A hole was drilled through the deckhead immediately above the manifold isolating valve, a spigot made to fit the handwheel and a bit and brace type removeable handle is kept on the helm- in the event of an emergency involving any of the systems off the manifold the helmsman can shut off steam in around 10 seconds by inserting the handle into the hole in the deck, through the engineroom deckhead, into the spigot and rotating the handle.

Greg

[SL Ethel]

Sorry for being daft, but can I ask for yet a bit more explanation as to why the two-piece bonnet is safer? From Wes' excellent pictures, it looks like with either design it would be possible to unscrew the bonnet when you merely intended to open the valve. How does the two-piece design prevent this?

Cheers,
Scott

[fredrosse]

Admittedly a rather long winded, but relevant valve explanation, SCREWED BONNET vs. UNION BONNET VALVES:

Either valve type could potentially unscrew its bonnet-body connection if one was to force open the valve handwheel, but it is much less likely with the union type bonnet arrangement.

The Union Bonnet type of valve has the threads that attach the bonnet to the valve body isolated from the working fluid (in this case steam and/or hot water), so those threads are less likely to potentially become damaged by fluid corrosion. Applying a turning torque to the valve handwheel directly applies equal torque to the screwed bonnet type valve threads. Applying a turning torque to the valve handwheel on the union type bonnet applies less torque to the union bonnet threads, as the joint between bonnet and valve body resists this torque, in addition to the bonnet threads also resisting the turning torque. Also note that the threads diameter on the union bonnet are much larger diameter than on the screwed type bonnet attachment. Larger diameter of the union threads, plus the resistance of bonnet rotation due to the bonnet/body interface, means that much more torque applied to the valve handwheel would be required to unscrew the union bonnet threads, hence less likely to accidentally unscrew them.

If the turning of the handwheel starts the bonnet threads turning, the union bonnet will allow separation of the bonnet-body pressure boundary, and this valve joint will leak profusely before the union bonnet threads are unscrewed any significant amount. Thus not allowing the bonnet to be blown off. For the screwed bonnet type valve, especially if there is thread sealant on the bonnet-body threads joint, the unscrewing of these threads may show very little leakage until the threads are run completely out, just then the bonnet is blown off.

Some union bonnets have a pin/ key or other geometry to positively prohibit rotation of the bonnet until the union threads are unscrewed enough to move the bonnet relatively far away from its seat. This type of union bonnet will positively prohibit bonnet rotation due to excessive turning torque on the valve handwheel.

[87gn@tahoe]

Thanks for chiming in with an excellent explanation Fred.

Sorry for the thread jack Josh