When Did Boiler Safety Get Serious?

[LMS2968]

This book is supposed to list all loco boiler explosions in Britain but misses out L M S 6399 Fury, witch makes one wonder what else is missed out

6399 wasn't a boiler explosion but a tube failure. It was one of the very high pressure tubes used to heat the high pressure part of the boiler, It wasn't part of the steam circuit, i.e. the steam within (from distilled water) was not used to power the cylinders. In these circumstances, it doesn't fit the definition of 'boiler explosion'.

[PenrithBeacon]

I think this is true. Henry Fowler wrote a report on this matter which was read into the LMS minutes. IIRC from the only time that I have read this minute book, the failure happened with a tube the design of which had been changed from the German prototype in order to fit it into the LMS loading gauge. The design change hadn't been tested, it being thought that the change was essentially low risk. How wrong they were!

 

Regards

[bertiedog]

As an aside there is some compression with water as air and gases are soluble in many cases, look at lemonade for example. Plain water will have a gas content, but very small, and a small pre-pressure and release causes it to bubble and surface. Of greater concern is hidden trapped air in larger boilers, this can be a bit dangerous to blowing off  leaking inspection covers, instead of a dribble, a bang.

All boilers are pumped up to several times( by formula) working pressure in several goes, with the whole thing repeated, then on to steam testing at working pressure.

 

Small boilers like garden railway are done like that as well, especially home built as the copper will still be in an annealed soft state from the silver brazing heat. The first pumping is done gently as the copper will stretch, and no violent loss should be applied as well. This ensures no unneeded strains, but after several pumpings the copper begins to harden off.

 

Tiny leaks can be peaned over to close, but best practice is to re solder. In 3.5 and 5inch leaks that develop after the test can be sometimes left to "take up" under close scrutiny. Soft solder caulking can be done, but it ruins the chances of a silver brazed repair later on.

 

The vital thing for garden rail is to get the safety valve working correctly, to go off at working, and also not dribble when closed which can ruin the performance.

 

The biggest safety factor for a small boiler is not to have more fuel than needed to boil the boiler dry, the gas must run out before the water gets too low. Most garden rail boilers are well designed and use tube boilers where the tubes are low in the barrel, or just use one main tube at the bottom.

 

The pressures are quite low rarely more than 30/40lbs, and a well made boiler will last years without any wear. a bit different from 225lbs, steel, and coal, plus the vibration and stress the real loco take.

 

Stephen

[jjb1970]

The incompressibility of water is why pressure testing is done hydraulically. Should there be a failure then if you squeeze hydraulically then the pressure drops almost immediately if there is a failure whereas if you do the test with air you can have a situation with a dangerous loss of containment of stored energy which can take some time to dissipate during which time the failure that caused a loss of containment could get much worse. Another issue which is more controversial is hot testing vs. cold testing and the pressure correction if doing a cold squeeze (you generally don't go as high if you do it cold). These days if doing a safety valve test it is more normal to do a trevitest type simulation rather than lifting the pressure to open the valve.

[New Haven Neil]

IIRC the Australians don't like 'model' boilers with stayed flat firebox sides - as in they are illegal?  The term Briggs Boiler comes to mind, is that a tubed firewall type of firebox - I need to go and research it!

 

Somewhere in interweb world I have seen a test to destruction of a garden railway type boiler, (I take an interest as I sell them for a living....) and it took well over 400 psi before the shell failed - not a joint.  They're pretty safe. I doubt the butane fire could generate enough heat to reach that temperature/pressure in practice.  I'm less convinced about the coal fired dry back boilers that people make in the garden scales, and have read in the press a terrifying tale which appeared to demonstrate the boiler was all but dry when feed water was added.  The person in question should have known better too. 

[Arthur]

 

Just for the book the volume of steam is 1600 times greater than the equivalent volume of water and so creates a very big bang when something fails.

 

That phenomena, the change in volume from water to steam, has been the cause of a number of ladle and furnace explosions in the iron and steel industry.

 

One of the worst being the explosion in 1975 at the Queen Victoria furnace, Appleby Frodingham, Scunthorpe.

 

To keep it simple, a large volume of water had entered a torpedo ladle containing around 100t of liquid iron. A crust had formed on the iron partially insulating the iron from the water. When a locomotive jerked the ladle into motion, a wave was created in the iron breaking the crust and several hundred gallons of water came into contact with 100t of liquid iron at around 1600 degrees C. The instantaneous boiling of the water created an explosive generation of steam which was unable to vent through the small ladle orifice.

 

Result, the ladle top, a heavy steel fabrication lined with refractory brickwork was blown clean off, the cast house was all but destroyed, every surface covered in a thin film of iron, fused slag and iron covered the floor and the heavy runner spout was found embedded in the roof framework. The ladle and 90t locomotive, a Hunslet Bo-Bo, were derailed and pushed back.

 

Seven men died and several were seriously injured.

 

.

[96701]

Didn't BR have its own boiler inspectors? Wasn't this because as a nationalised industry it didn't have any insurance, just met any claims out of its own pocket. BT was the same until it was privatised.

 

Andy G

I remember whilst working in the Regional M&EE office at York with the lift inspectors, lifting machine inspectors, electrical installation inspectors and boiler inspectors, Wilf Gargett put his coat and flat cap on and stated that he was about to go to the museum and give the Duchess an internal.

[PatB]

Just for the book the volume of steam is 1600 times greater than the equivalent volume of water and so creates a very big bang when something fails.

 

And, of course, what makes a boiler explosion really dangerous is the fact that, as the boiler fails, the pressure inside drops suddenly and all the liquid contents pretty much instantly flash into steam with the consequent increase in volume .

[PatB]

 

That phenomena, the change in volume from water to steam, has been the cause of a number of ladle and furnace explosions in the iron and steel industry.

 

One of the worst being the explosion in 1975 at the Queen Victoria furnace, Appleby Frodingham, Scunthorpe.

 

To keep it simple, a large volume of water had entered a torpedo ladle containing around 100t of liquid iron. A crust had formed on the iron partially insulating the iron from the water. When a locomotive jerked the ladle into motion, a wave was created in the iron breaking the crust and several hundred gallons of water came into contact with 100t of liquid iron at around 1600 degrees C. The instantaneous boiling of the water created an explosive generation of steam which was unable to vent through the small ladle orifice.

 

Result, the ladle top, a heavy steel fabrication lined with refractory brickwork was blown clean off, the cast house was all but destroyed, every surface covered in a thin film of iron, fused slag and iron covered the floor and the heavy runner spout was found embedded in the roof framework. The ladle and 90t locomotive, a Hunslet Bo-Bo, were derailed and pushed back.

 

Seven men died and several were seriously injured.

 

.

 

I heard about this one because my first boss, when I came to Australia, had been present on site (though, fortunately, well away from ground zero) at the time.

[PatB]

IIRC the Australians don't like 'model' boilers with stayed flat firebox sides - as in they are illegal?  The term Briggs Boiler comes to mind, is that a tubed firewall type of firebox - I need to go and research it!

 

Somewhere in interweb world I have seen a test to destruction of a garden railway type boiler, (I take an interest as I sell them for a living....) and it took well over 400 psi before the shell failed - not a joint.  They're pretty safe. I doubt the butane fire could generate enough heat to reach that temperature/pressure in practice.  I'm less convinced about the coal fired dry back boilers that people make in the garden scales, and have read in the press a terrifying tale which appeared to demonstrate the boiler was all but dry when feed water was added.  The person in question should have known better too. 

 

Model and miniature boiler design in Australia is governed by a series of codes. One for steel boilers, one for copper and one for "sub-miniature" boilers (effectively the garden gauges. The only one I've read in detail is the sub-miniature volume, which details ultra conservative design standards and working pressures (not more than 35 psig). As a result I can't comment on the local view on stayed fireboxes.

 

The Briggs boiler is a dry firebox design. The best simple description I can give is of a cuboidal firebox set into the rear and lower half of a cylindrical boiler. The only surfaces of the firebox in contact with the boiler are the crown (flat in the designs I've seen) and the upper portion of the tubeplate. I've always assumed that the design came about for simplicity of construction rather than any inherent safety advantage, although I would think that adequate inspection would be comparatively easy too.

[ahardy]

This book is supposed to list all loco boiler explosions in Britain but misses out L M S 6399 Fury, witch makes one wonder what else is missed out

Interestingly I think this book misses out the first instance of a locomotive boiler exploding. This occurred on the Middleton Railway in something like 1814 or 1815.

 

Andy

[34theletterbetweenB&D]

Hewison has it that 'Brunton's Mechanical Traveller' 31/71815 at Philadelphia was probably the earliest, and worst in the UK in respect of lives lost, 16 killed.

 

That said, both in the opening of the first chapter and subsequently he makes it plain that records of early UK locomotive boiler explosions are very incomplete and unreliable. On p29 is the marvellous description of boilers exploded in 1845, possibly one engine named  Irk or Erck, or possibly they were two different engines; but whatever, a Railway Inspectorate officer Lt-Col Wynne writing in 1857 referred to one of these as 'one of the earliest'. That's all the evidence I need that record keeping was unreliable at the time!

 

What is also clear is that until relatively late if there was no loss of life or damage outside the locomotive operator's property, then it was perfectly normal not to report the event. The significant 'markers' that Hewison picks out are 1871 when any event 'of a kind as to have caused or be likely to have caused loss of life or personal injury' should be reported - notice that there remains ample wiggle room in that phrasing, for anything that happened 'out of sight' on private property to go unreported - only closed off in 1895 by a Statutory Order defining a burst boiler as a reportable incident.

 

This book is supposed to list all loco boiler explosions in Britain but misses out L M S 6399 Fury, witch makes one wonder what else is missed out

 Tube failures were a commonplace, so they are all missing. Arguably properly so, as they do not cause the main pressure vessel to fail, which is the inevitably dangerous and destructive event; even though there is probabilty of injury and risk to life, principally of footplate crew.

 

I imagine that a tube failure in an operating locomotive boiler will be a reportable accident by now?

[petethemole]

It assume it would be a Reportable Dangerous Occurrence as shown here under 'Pressure Systems' http://www.legislation.gov.uk/uksi/2013/1471/schedule/2/made

 

"Pressure systems

2.  The failure of any closed vessel or of any associated pipework (other than a pipeline) forming part of a pressure system as defined by regulation 2(1) of the Pressure Systems Safety Regulations 2000(1), where that failure could cause the death of any person."

[Nearholmer]

During an exceedingly rare bout of ironing this morning, I was pondering the relative dangers of modern steam-generator irons, where steam is generated in a 'docking station' and conveyed at 6bar (I must check that, it seems pretty high) to the iron via a flexible tube, and 16mm/ft live locos.

 

Using the bar.litre measure, even if the steam reservoir for the iron is very small, it probably contains more stored energy than a loco boiler.

 

As far as I know, no SMEE routine inspection and test regimes are applied to irons, and I have to say that the flexible tube looks like a bit of a hazard site to me.

 

Kevin

 

[That figure of 6bar is correct, and some top-notch domestic irons deliver up to 7.5bar. 16mm/ft locos typically operate at 2, 3, or, exceptionally, 4bar, IIRC.]

[Reorte]

During an exceedingly rare bout of ironing this morning, I was pondering the relative dangers of modern steam-generator irons, where steam is generated in a 'docking station' and conveyed at 6bar (I must check that, it seems pretty high) to the iron via a flexible tube, and 16mm/ft live locos.

 

Using the bar.litre measure, even if the steam reservoir for the iron is very small, it probably contains more stored energy than a loco boiler.

 

As far as I know, no SMEE routine inspection and test regimes are applied to irons, and I have to say that the flexible tube looks like a bit of a hazard site to me.

Thanks for that - now I can say I'm playing it safe, rather than being too lazy, for hardly ever ironing.

[kevinlms]

Model and miniature boiler design in Australia is governed by a series of codes. One for steel boilers, one for copper and one for "sub-miniature" boilers (effectively the garden gauges. The only one I've read in detail is the sub-miniature volume, which details ultra conservative design standards and working pressures (not more than 35 psig). As a result I can't comment on the local view on stayed fireboxes.

 

The Briggs boiler is a dry firebox design. The best simple description I can give is of a cuboidal firebox set into the rear and lower half of a cylindrical boiler. The only surfaces of the firebox in contact with the boiler are the crown (flat in the designs I've seen) and the upper portion of the tubeplate. I've always assumed that the design came about for simplicity of construction rather than any inherent safety advantage, although I would think that adequate inspection would be comparatively easy too.

The Australian code history can be found here.

 

http://www.aals.asn.au/AMBSC/AMBSC.htm

 

Edit to add shop details

 

http://www.smex.net.au/Store/Store_AALS-Codes.php

 

I believe the Code Part 2 Section 8, makes it next to impossible to get imported loco boilers passed. I can remember the late Doug Baxter (the last surviving Victorian Railways boiler inspector by a considerable margin. Others had seen the writing on the wall and transferred to Harris suburban train construction - sadly, this was the wrong move, as they contained large quantities of asbestos), stating that several people had imported locos & there was no way they could ever be made to comply with Australian standards. A new boiler being the only option.

 

Doug assisted all sorts of steam projects and was highly sort after and hadn't forgotten a thing!

 

http://home.mysoul.com.au/paul.hallett/r_711.htm

[Siberian Snooper]

During an exceedingly rare bout of ironing this morning, I was pondering the relative dangers of modern steam-generator irons, where steam is generated in a 'docking station' and conveyed at 6bar (I must check that, it seems pretty high) to the iron via a flexible tube, and 16mm/ft live locos.

Using the bar.litre measure, even if the steam reservoir for the iron is very small, it probably contains more stored energy than a loco boiler.

As far as I know, no SMEE routine inspection and test regimes are applied to irons, and I have to say that the flexible tube looks like a bit of a hazard site to me.

Kevin

[That figure of 6bar is correct, and some top-notch domestic irons deliver up to 7.5bar. 16mm/ft locos typically operate at 2, 3, or, exceptionally, 4bar, IIRC.]

Where's the I'm flabergasted button, when you need it, that's 80 odd psi too over a 100 psi. I'm glad I missed the ironing course.

[Nearholmer]

Yes, I still sort of don't believe it.

 

If you want to see one in pieces, this chap has done it for you https://www.ifixit.com/Teardown/Philips+Pressurised+Steam+Generator+GC8220+Teardown/26948 and here is a repair manual https://www.chsinteractive.co.uk/pdf_data/DIA00429.pdf

 

If you look at the steam valve, you will see that it is a serious piece of engineering.

 

They seem to be fitted with a sort of fusible plug thing, a one shot safety valve, which discharges steam downwards. The manual for ours says that if steam starts to come out from under the unit, it must be switched off, and returned to the manufacturer (after it has finished discharging steam and hot water all over your house, I guess). [edited to remove error of fact]

 

If there is an expert on these irons here present, I would love to understand the safety features better.

 

Kevin

[jjb1970]

A few years ago there was a period where getting hold of a pressure vessel surveyor from the insurance companies was painful. One of those coffee boiler things in a supermarket coffee shop blew up and a bit of the debris caused a very nasty injury to a customer which resulted in amputation of a limb I believe. Somebody realised that the things should have been considered to be pressure vessels and subject to inspection and that they weren't so there was a crash program to inspect thousands and thousands of the things. I was in a power plant with 260Bar+ boilers at 565C steam temp and we couldn't get insurance surveyors because they were too buying inspecting coffee machines.

[New Haven Neil]

Nearholmer - I wouldn't say 4 bar is exceptional - all Accucraft locos blow off at 60psi - even Aussie market ones I believe.

[Nearholmer]

I'm sure you're right; I couldn't remember for sure, which is why I caveated IIRC.

 

I only, very briefly, owned one Accucraft loco, a Porter 2-4-2 tender engine, which ran very well, and was exceedingly pretty, but was really bought as speculation, when I spotted it for sale ludicrously cheap!

 

K

[bertiedog]

Nearholmer - I wouldn't say 4 bar is exceptional - all Accucraft locos blow off at 60psi - even Aussie market ones I believe.

That's a bit high for garden rail type locos, quite safe, as it is the safety pressure, but it indicates the cylinders and valve gear are not really set for efficiency. I suspect they are using the pressure to over come a tight system, not worn in any way. With steam models a lazy slightly worn system with good valve setting can be more efficient than a tighter non leaking engine.

 

Most model engines run on gas, so to regulate the overall pressure is easy, when the safety valve blows the gas can be turned back to the running pressure via the gauge.

 

The pressure applied to the piston is far lower than the boiler pressure due to heat loss in the pipes and castings, and the pistons. Model locos are rarely lagged very well, and have rather long steam pipes.

 

All garden rail benefit from a long run on air to run them in properly, with air line lubrication of course, this will transform the performance before steam and the vital component, proper steam oil, is used.

 

If you find that a particular engine blows at 60, but you have to cut it back with the gas during running, it really needs a lower set safety valve replaced on it.

 

A problem with overall performance on the simpler types of boiler in 16mm is the non shut off of the valve, even a small leak and dribble is throwing away large amounts of power. A good valve snaps open and snaps closed, but they are not easy to make! They should also open and close within a tight tolerance, too wide and huge amounts of potential power are wasted.

 

On the steam iron safety, it must depend on the potential stored energy in the system and I suspect it is far lower than it appears at first glance. The heater would be designed to be "lossy", little insulation and modest volume of metal storing heat. The flow of water controls the potential heat, and in idle the power is shut down, only heating again on demand. They seem safe, the stuff escaping is not steam but condensed steam at a much lower temperature.

 

Measuring the temperature would reveal the pressure at the nozzle. The boiling container would be higher , but I suspect that just like any sales talk the pressure may not be as high in operation as the sales brochure suggests. The test pressures would be very high, and maybe the design keeps heat in the boiler to maintain the high pressure when idle, but then it drops dramatically as the steam vents to atmosphere.

 

The whole design sound dramatic, but operates in a way towards safety at all times.Some of the claims to pressure on water car washers etc are very miss leading, as so little water is actually under the high pressure at any one point.

 

Stephen

[Nearholmer]

Good stuff, Bertiedog.

 

I think you are right about the irons, because the quoted pressure is at the boiler. Once the steam valve opens, steam is being delivered down a long tube, to a large number of outlets.

 

If the iron was simply pointed into free air, the pressure gradient would be from, say, 6bar at the steam valve in the base unit, to 1bar at the outlets.

 

With the iron pressed down onto a shirt on an ironing board, a high proportion of the pressure drop should occur through the shirt and the padding on the ironing board.

 

And, my most effective 16mm/ft loco, one of the two I kept when selling-up to go indoor 0 scale, is almost certainly the least efficient. Internal gas fired, but feeding steam via terribly long pipes to oscillating cylinders, within which the fit of the pistons is anything but tight. By careful fiddling with the gas control, reverser (which acts as a regulator, if not moved fully across), and the needle-valve regulator, it runs on very low pressure, damp steam. It plods up hill and down dale at a constant speed, will take any sensible-length train hung behind it, and can be kept going like that "forever" by topping the boiler and gas up periodically. It throws away most of its lube oil in the first five minutes, but since it is running on near-condensate, that isn't a problem. No firm idea what pressure it runs at, no gauge, but my gut feel is 1.5bar, and the hardest part is nipping the gas valve down far enough to prevent it making too much steam and lifting the safety valve (3bar), without accidentally turning the gas off altogether. It is the perfect example of what you term a "lazy system".

 

Kevin

[peach james]

That's a bit high for garden rail type locos, quite safe, as it is the safety pressure, but it indicates the cylinders and valve gear are not really set for efficiency. I suspect they are using the pressure to over come a tight system, not worn in any way. With steam models a lazy slightly worn system with good valve setting can be more efficient than a tighter non leaking engine.

 

 

 

Stephen

 

 

?

 

Really.  I have a Reeve's Monarch look-a-like that runs quite nicely at 30 lb, and a Stuart V-10 that is set for either 80 or 100 lb/inch ^2.  Why?  Because one is more loaded than the other- and I actually need some reserve with the Stuart, not like with the Monarch.  (which originally had a 10V on it way back when I was 11 or so...).  The relative tightness of a commercially made engine is not going to require the 60 lb safety setting, but it makes it far easier to have the margin between maximum needed for scale operation (probably as low as 20 PSI), and a sensible safety valve setting.

 

And n for a small steam engine?  Who cares...now, does it do the job, that does matter, but actual n is a totally non-sensical calculation or value.  Look at IMLEC values, and they have historically been <2% n or less (and worse with 3.5" gauge...).  The valve gear settings which give most "efficient" operation are not those which are commonly used on model engines.  

 

Britcar1 by Peach James, on Flickr

 

was taken on the Thursday before OMLET 2000- when I soundly trounced dad (he was on Caribou, I was on Britannia) by 2:1.  Both are effective engines, but one was being operated far more efficiently.  It still would have been much more efficient to leave it sitting at home...

 

A huge amount of this can be figured out from reading "The Generating Game" in MRC (right to the bitter end, June 1986)- which went through the strength of a boiler fairly easily, and showed how strong the average pot boiler actually is.  I did a very brief set of calculations earlier in the week for someone asking questions, and to give some idea, a steel boiler 12" diameter could be as thin as .020 at 100 PSI before failure.  (no safety margin !).  (the calculated was .200 at 100 PSI, and 10x safety factor, hoop stress only)

 

There is, for practical purposes, not a lot of difference in energy in the boiler between 20 PSI and 60 PSI.  (roughly, 20 BTU/Lb + the metal), which is the best part of nothing compared with the 1000 BTU that the steam already has in it.  So, it makes control a little easier to have the higher pressure, giving a little bit more wiggle between maximum haulage capacity and safety valves lifting.  (close the damper, open the door, put the pump on).  I would leave any system with a 60 psi safety valve alone as regards boiler pressure, and just throttle in on the fuel supply if the engine operates better at 40 PSI or less.

 

My Britannia is set the same sort of way- the safety is set at 80 PSI, with much above 40 psi being basically "spare", because the wheels slip at full regulator/reverser at about 40 PSI.

 

My experience is more a bit bigger- 3.5" gauge live steam, and 1 1/2"-2" road steam, for the most part.  However, this thread has gone a long way towards fear mongering, IMO, compared with what it started with.

 

While I cannot speak for BR, I know the Royal Canadian Navy had it's own boiler inspectors, (and may still on the west coast, although I understand that Barry has retired).  I would expect that BR would have done the same thing, and been self insured as an arm of the government. 

 

100_0092 by Peach James, on Flickr

 

James Powell, 3A (PRO)