I'd make the point that in such an event if you were on the engine crew (particularly the fireman) you'll be in for some pretty intense questioning by authorities as to why you ran when there may have been a way to recover or reduce the severity of the situation- at the very least as Rick mentions turning the injector on and attempting to stifle/turn down the fire before bailing. If they ask why you DIDN'T put water in the boiler, you'd better be ready for a complete defense of why you believed it would cause the crown to rupture or would otherwise have a negative effect- just sayin'.

At any rate, if you look in the firebox to see that the sheet is visibly glowing- it's likely already lost enough structural integrity to be within seconds of failing (I don't have the table immediately on hand for how much tensile strength is lost as temperature progressively increase), not enough time for anything you do to matter much anyways.

Let's hope to God none of us are careless enough to find ourselves in this type of situation- a fatal BLEVE from a locomotive will surely spell the end of much of what we do.

What I was taught was to immediately drop the fire, or as much of it as might not be needed to prevent thermal stresses from cold air. Radiant heat should take the crown heat down at a controllable rate. Quench it unevenly with effectively cold water and things may take a much more prompt course.

I keep thinking about a detail of the N&W Class A wreck described in Ed King's book. An A was derailed nose-down, coming safely to rest but obviously with the crown exposed. After some time, the engineer went down to cut on an injector, at which point the boiler promptly exploded with tremendous effect.

Keep in mind that the crown will only tear or sag enough to vent the mass flow of evolving water and steam. That may not be the full catastrophic failure that might occur with the engine being fired under load.

Here’s another story and approach for dealing with low water.

My dad was almost witness to a boiler explosion due to low water at a local sawmill when he was a kid in rural SC in the late 1920’s. He was awakened in the middle of the night by a steam whistle continuously blowing which he recognized as coming from the nearby small sawmill, about a mile from where he lived. He jumped up, threw on some clothes, and took off running towards the sawmill to see what was up. Halfway there, my dad met the fireman running away from the sawmill yelling “it’s gonna blow up!”. Despite the fact he was completely winded, my dad had to immediately turn around and run back the way he’d come!

Apparently the fireman had fallen asleep on the job, allowing the boiler water level to get really low. He pulled the whistle cord and tied it open in an attempt to lower the boiler pressure and took off running. Apparently the fireman was successful, as the boiler didn’t explode.

This story doesn’t give a lot to go on, like how low the water level actually was, what was the state of the fire, or did the fireman start an injector (I expect not). I wonder how rapidly a fully open whistle would lower the boiler pressure? Perhaps in a small operation like this one, the whistle might have been large relative to the size of the boiler so it could drop the pressure reasonably quickly.

[quote="Rick Rowlands" Either way you can still run away from the locomotive. It only takes seconds to turn on the injector.[/quote]

Or, you might not need to run having been blown away already. Seconds matter.

There is a big difference between cresting an hill and watching the water vanish, then coming up on a low water situation that for some reason, no one knows how long, how low, how bad it is.....

The soft plugs are not there to vent pressure at the last second. Pure tin plug will melt at about 450 degrees. A 250psi boiler has water at 400 degrees. You're just barely out of water and the crownsheet would be just fine when the plug blows, which is the point of a safety device. Kind of like a circuit breaker on something electric...you don't want it to trip when the wires are already on fire, you want it to trip before any additional damage happens.

On the blowdown, every gallon of water vented will remove about 3 million joules of energy(from the above 250ps boiler) from the boiler. It won't drop a lot of pressure or cool the boiler much until you are close to out of water, but you are pretty rapidly removing energy from the system

As for anyone who's main concern is that they are going to end up with a dead engine needing a tow, you scare me. That kind of attitude gets people killed on probably a daily basis, in industry.

Agree 100%. It's my opinion that if an engine is ever suspect of having incurred a low water event (Not just momentary "sloshes" typical of normal operation) it should immediately be taken from service and shut down asap until qualified persons can inspect the boiler and make proper judgements.

The consequences are too high for any other approach. Having an engine OOS and getting a tow is a better outcome than pushing your luck and ending up with far, far worse.

Mr Pegasuspinto & Mr Boilermaker,

Sorry if my not wanting to tie up a nationwide common carrier scares you. I've directly prevented a DOZEN deaths during my railroad career, and an unknown number of injuries because of my concern for safety. Feel free to check with anybody who has ever worked with me, any of my previous employers, or the FRA if you doubt my word.

I notice two things about this thread:

First, there seems to be a number of divergent opinions about how and why an overheated boiler will fail, or how to handle it to prevent such an incident. (As an aside, the suggestion of just adding water to a dry boiler to top it off scares ME.) I'd say it's true that (so far as I know) we don't have detailed, specific data covering these type incidents. As Mr Ellsworth pointed out, it's kind of a crap shoot.

Second, we seem to be in agreement that there is only one way to be sure that a boiler won't explode. KEEP UP WITH YOUR WATER LEVEL.

Any questions ??

This debate continues without any real answers. I agree, that there's a lot of old wives tales and very little solid evidence of an incident when adding water made the crown sheet crack.

The best theory I've seen for why that's such a common tale is from cast iron being overheated and cracking when quenched. Hopefully nobody is using a cast iron boiler.

My personal approach would be quite simple. Turn on the injector. If you're an experience fireman, that can be done very quickly. As you do that, which should almost be instinctive, if you truly believe the crown sheet is exposed and you're in danger, advise everyone is the cab to vacate immediately, be certain they do, and then do so yourself. Observe the experiment from a safe distance.

As a side note, we used to tell cab riders that we would never make any jokes about something like that or try to scare them. "If we tell you to leave the cab, DO IT. Don't ask questions, don't hesitate." The intent was for something like a busted water glass or a lubricator that starts spouting oil, not the crown sheet but it works for any dangerous situation.


Should you have a loaded passenger car adjacent to the locomotive, then your choices get far more challenging. Do you try and cut away? Best to not get there in the first place.

I wonder if we shouldn't establish a standardized term similar to the phrase used in military aircraft -- repeat some simple phrase like EJECT three times when it's time to go without hesitation.

The original point of this thread involved forces that complicated any 'de-heating' of an overheated crown sheet, possibly through softening parts of it to failure. If that condition is established, whether or not the injector is on, you're likely to proceed to failure, perhaps even if you promptly drop the fire and the plate starts to cool from the fireside with the legs full and 'dark'. There just can't be any one sure or safe outcome if you were to find yourself in this situation.

As Sammy points out -- be diligent about not getting near the state in which Eisenhoffer and Leidenfrost reveal themselves to be Murphy and Finagle in drag.

I will start by saying most of my knowledge is based on my own experiences running live steam locomotives so I am not sure how well it will translate to the big ones, none the less how do hills factor into all of this?

Say you're firing 4014 going up a hill with water rather high in the glass you come over the top and watch your glass empty out as the water sloshes forward. I think yelling eject and expecting you and the rest of the cab to go tumbling out the cab at 65mph seems like a bad idea!

With that being said a slosh is one thing but how long is too long with no water in the glass? Even if you lose it at the top of the hill is it really gone or more than likely covering most of the crown sheet if not the whole thing and just tapering off right in front of the glass? I would think hitting the injectors would be the no brainier for low water coming down a hill. On a larger engine like say a 4014 where should the water be at the start of a large grade 3/4's glass, and how far would the fireman expect it to drop when he comes back down the hill before he should start getting nervous?

The other one that makes me wonder on Strasburg's 475 the boiler extends all the way back in the cab and the water glasses are set all the way at the front of the boiler. Do they have those set do that they're level top the top of the crown sheet? If so and the boiler is pitched down it seems like you could be seeing a decent reading in the glass but still have some crown sheet uncovered at the front of the boiler. Is that an issue?

I can tell you that 4014, and no doubt many other big engines, has two sets of water glasses stacked on top of each other. What they show at the bottom of a hill I do not know, but you can bet that they have the water going before they top over any kind of summit.

Federal regulations require that the lowest possible reading of the sight glass shall not be less than 3 inches above the highest point of the crown sheet- the same applies to tri-cocks. Regardless, it's all about familiarity with your locomotive and the territory you are running on to judge "false high" or "false low" readings due to grades, and prepare for when you know you will be accending, cresting, and descending a hill.

Just because the water is gone from the glass doesn't mean that it is gone from the top of the crown sheet. There still is 3" of room there and in the location where the greatest number of gallons per inch is located. More than likely the water is somewhere in those three inches of space if your water disappears from the glass but you knew you had water a few minutes ago and you haven't done anything that would use up a large amount of water.

Since our railroad is basically two hills, one up and one down with a level tail track, we make it a habit of checking water level while on the level part and getting it to the right location there. When on the uphill the water glass is practically useless and on the downhill having only an inch of water in the glass is actually the right level so as to not be overfull when we go back up the hill. It is all part of how we have learned to operate on this railroad.

People ask me when I am going to put a reverse loop at the end of the railroad and the answer is never as there will never be a way for the steam locomotive to ever face downhill on our 6% grade. Nope, never, ain't going to happen!

It's been mentioned before, but Dixie 576 had a 'plate' on the water glass indicating correct level for a nominal 2% grade, to simplify management.

Two hundered years after 1756 when Leidenfrost described his "phenomenon", it has been studied and quantified and in a sense made generic as it is described now as Departure from Nucleate Boiling (DNB) heat flux (i.e. heat transfer rate). How much heat can you shove through a surface with water on one side and the boiling changes from nucleate boiling (small bubble) to film boiling. It is one of the most critical values derived in my "How to Design a Nuclear Reactor in 13 Easy Chapters" college textbook. It determines how much heat you can remove from a reactor core without causing a meltdown. Which is pretty much the same as running a dry crownsheet.



I have been in all of the boilers of every steam locomotive in Hawaii over the last 40 years and was good friends (from when I was 13 yo) with Jay Conde, author of all the Sugar Train books and noted historian. Over 250 steam locos were in Hawaii and in all of history not one explosion was recorded. Not that they did not get low watered or mud burned, but no booms. I am not as familiar with Rocky Mountain narrow gauge, but no big booms come to mind.

I have read an article from the 1880s, about a PRR test in the late 1860s, to test the water on a hot crownsheet boom theory. The author could not locate the documented results from PRR but had access to the industry journals describing the test. They took a serviceable boiler, put it in a field, built bunkers to witness the results and tried to blow the boiler up. They couldn't do it. Four or five attempts and just got leaks. Fixed the leaks, fired it again, low watered it, raised the water, ad infinitum......

The difference between a crownsheet failure (Gettysburg) and an explosion depends on the rate of pressure release above the waterline.

Mechanics of an Explosion

About 20 years ago, I did some calcs and basically determined that flashing water to steam on a hot crownsheet would only trigger an explosion if the crownsheet was about 15 seconds from launching anyway.

Here are the scratch pad calcs:

Assume a K-36 with 4' x 4' of the crown red hot. Thickness is 3/8"
Under high fire, the sheet will be about 700F, take away the water, the temp climbs to orange color about 1200F. A 500F difference. total BTU increase 13,700.
How much water will that boil? About 17.6 lbs of water or about 2 gallons. How big of a pressure spike while all of this occurring at 200 psi in a boiler generating about 20,000 lbs of steam per hour?

The real kicker that quenches the "cause a boom" theory is that the steam production rate is SEVERELY REDUCED because the steam generation is occurring above the Leidenfrost/DNB point.

So, to water a dry crownsheet or not? How close are you to the 15 to 30 second window of inevitability anyway? To quote my favorite Dirty Harry line, "Do you feel lucky???"