Conditions inside a boiler

As a spin-off from the Flying Scotsman thread, I came across these videos that filmed the conditions inside a boiler in operation (via a small viewing port). The actual boiler is stationary plant, rather than a locomotive boiler, but there is a lot of interest nonetheless, particularly with regard:

- The appearance of the surface of the water
- What happens to water level when demand is increased
- What happens when water is injected
- The effect of increasing total dissolved salt concentration

Best viewed in sequence, as each one builds on the previous.

1 - Introduction



2 - Basic boiler ccontrol



3 - Water level and gauges



4 - Feedwater control



5 - Low pressure operation



6 - Increased demand



7 - Very high demand



8 - TDS control



9 - Summary



Tom

 

Spirax Sarco produce a huge range of training aids on steam and boilers, available on the Net and well wort investigating if you have the time. As you say a lot of it is not directly relevant to steam loco boilers but it does help you better understand what is going on. There is another video which shows the take off from a boiler and the effect of increasing demand. I think I've referenced it on Nat Pres before but it was a good while ago. The water level forms a volcano under the take off and is significantly higher than the surrounding level at this point. People often say that opening the regulator causes the water level to rise in the boiler, which it does locally but the Spirax video quite rightly says that this is not observed in the gauge glass. So, for your homework, why does the level change when the regulator is opened?
I've had first hand experience of what can happen with a boiler when the demand is suddenly and significantly increased. A waste heat boiler (and so akin to a loco boiler where you can't instantly turn of the heat), it ran at 10 bar during the day. it was connected to a hospital and the 10 bar was necessary for the laundry (don't ask me why!) However, when the laundry shut down at 18.00 each day the system pressure was lowered to 6 bar (again don't ask me why!) As the hospital demand was still largely there for heating the was still a significant take off of steam. However, the control system now said the boiler pressure was too high and brought condensers on line to get rid of the excess steam. These condensers are rated at full boiler power so the steam demand effectively went from about 80% demand to 180% demand in a few seconds. The result was a huge water loss well in excess of that which the pumps could meet and low water levels with pulled tubes as a consequence. Fortunately no explosion or major damage. The control system was modified to reduce the boiler pressure in steps but would have benefited from a control system that acted on an actuated pressure reducing valve to control system pressure and not by varying the boiler pressure.

 

That's an interesting question ... (and very obvious on our Chatham locos, where the rise is very marked).

Anyway, let's simplify a bit: We'll assume the loco is working on level track (so any change in level is not due to the inclination of the boiler) and working at constant speed (so it is not due to the inertia of the water causing surging). We'll also assume that the overall boiler pressure is constant while working (so the heat input is sufficient to balance the steam being used by the locomotive).

The regulator opens and the level immediately goes up in the glass.

If the level goes up in the glass, it must be because the apparent density of the water has gone down (so the same amount of water takes up more space). That can't be due to rising temperature (because the pressure has stayed constant). So my assumption is that it can only be because within the main body of water (below the foam), the space taken by bubbles has increased.

Because the pressure is constant, it can't be water suddenly flashing to steam; and likewise, the average bubble size shouldn't have changed (because the bubble size is related to the pressure). So it must be due to enhanced production of bubbles within the main body of water, which would be the case since (with pressure stable while working), the steam production rate must be higher than is needed to maintain a constant pressure with no demand.

So essentially, the loco is moving along level track with the regulator shut, with a given water level and the fire just maintaining an equilibrium pressure. The regulator is opened, draft goes up, fire burns more fiercely and the rate of steam production therefore increases. So more bubbles in the main body of water; the water density therefore goes down and the water rises in the glass.

Is that it?

Tom

 

The water in the glass is not bubbling so not subject to these changes in density.
For what reason have you opened the regulator? Think outside the boiler to answer this.

 

Going outside my comfort zone but I'll have a crack!

Perhaps as steam is drawn off when the regulator opens the pressure inside the boiler drops accordingly thus creating more space for the water to fill and therefor the water level rises. However, as I understand it, you can only compress water 4/5 of a bees dick, so unless more water is entering the boiler somehow the water may not suddenly fill the extra space just by itself allthough the physics of water under boiler conditions are way beyond anything my high school education had to say. Could it also be that as steam draws off and the pressure is reduced, the boiler itself contracts and the water level rises as the same volume of water is now contained in a reduced space?

 

Outside of the boiler implies some kind of mechanical effect.

Opening the regulator gives a force from the pistons, which is imparted through the whole structure, and since I am assuming constant speed and level track, that force imparted to the structure must just be sufficient to overcome the train resistance, i.e. friction etc.

If there is not a measured change in density, then the only other option to raise the water in the glass is if the water is higher at the back (where the glass is) and lower at the front (so the total volume remains constant).

If we think of the whole loco as a rigid mass, then the force applied acts on that whole mass. Assuming the boiler to be a tube part-filled with water, some of the force must be acting on the water, i.e. there must be a pressure differential between the backhead (pushing in the direction of travel) and front tube plate (resisting).

Since the water and steam are essentially a closed system, the steam can't exert that pressure differential. So the only thing I can think of is that there is a head of water, i.e. higher at the back and lower at the front. So even at equilibrium (constant speed, boiler pressure, gradient), the need for a force on the structure of the loco to overcome friction means the water would rise at the back and fall at the front (defined relative to direction of travel), giving an apparent rise in level. (Going backwards should result in a fall in the glass).

If that is right, wouldn't that suggest that even in a cold loco (but with water in the boiler) being towed by another, the water level would rise when moving?

Tom

 

That's pretty much as I believe it to be, Tom. Because you have opened the regulator you are applying a force to the loco, either to accelerate it or climb a gradient. The only way the force can be applied to the body of water is by a pressure difference. We are not talking high pressure One inch of water is only about 0.04 psi.
I tried dragging a loco up a gradient out of steam and the water level rose in the glass.
And yes, the water level falls when going in reverse uphill.
Because the water level changes when the reg is opened people associate it with the upsurge of water under the steam collection point but this is generally well away from the gauge glass and, as Spiral Sarco say, is not seen at the gauge glass. I used to believe it was due to the surge of water till I saw the videos and it set me thinking. An exception is if the steam take off is over the firebox close to the glasses (such as some Manning Wardle) and in this case the result is very noticeable as the water rises a good few inches. . GWR loco generally have the steam fountain right at the back of the fire box an the rise in water level is quit noticeable when you open the injector steam valve.
Award yourself a Mars Bar.

 

Metaphorically consumed ....

Thinking about this further. If you have two boilers of different shape containing equivalent mass of water, the force needing to be applied to that water under the same conditions of gradient and acceleration etc. should be the same. However, the pressure (and therefore head) required to generate that force will vary depending on the precise geometry.

Without going into calculations (which I can't do in my head ... ), for a given mass of water in the boiler needing to absorb a given force, to minimise the required head (i.e. water lift), I think you you want lots of area at the back and not much at the front

So, all else being equal, you should get less lift in:

- A Belpaire firebox, relative to a round-top box
- A taper boiler relative to a parallel one.

Which, in as much as anything can be said to be "all else being equal" seems to accord with my experience ("Camelot" with a Belpaire and taper boiler seems to lift less than the similar sized 847 with a round top box and less taper etc). Of course, it is complicated by the fact there is also a change in level due to gradient that is dependent on boiler length.

Tom

 

It is indeed complex, way beyond me and my slide rule to calculate it for a boiler. It's a law of physics so would hold for any shaped vessel. I think that the easiest way to consider it is to have an open topped vessel with a round bottom of constant radius. i.e. half a circle. If you part fill this vessel with water the surface at each end remains the same even if it is tilted on a gradient. Hopefully the attached sketch will be self explanatory. When the vessel is stationary, everything is in equilibrium and H1 = H2. If we impart a force on the vessel to accelerate it or climb a gradient the only way the force can be imparted to the water is by a pressure head. Thus, H1 must be greater than H2.

 

Whilst we are talking about water in gauge glasses, the water in the gauge glass is outside of the boiler and is not actually boiling, which is why you get a generally steady level. However, on many occasions I have observed the water in the gauge glasses of 60007 bubbling quick significantly, preventing you reading the water level. Initially, I put it down to foaming (as do others) because, if you shut the regulator, it stops. However, I've since come to the conclusion that it is simply boiling as, if you soak a rag in cold water and bathe the bottom water cock, it quickly stops until everything gets hot again. However, a couple of things confuse my simple mind because I can't think of a reason. It only ever happens when running in reverse and it is generally the left hand gauge glass that does it. The right hand will occasionally but only after the left hand one has been doing it for a while.
It is a phenomenon I have only ever come across on 60007. Has anyone any similar experience?

 

Not something I've ever experienced.

Where is the steam take off in the boiler, relative to the position of the gauge glass?

Tom

 

You may be over thinking this but I think your conclusion is correct. Taking @Steve 's example of an engine towed out of steam, the average surface of the water (ignoring surging) is going to be perpendicular to the total acceleration vector. If you take 1g truly upwards (as the acceleration in opposition to that which would be imposed by gravity if the ground wasn't there) then superimpose the longitudinal and vertical accelerations of the loco on that, the surface will be perpendicular to the resultant vector.

The effect at the gauge glass will be less marked on a loco with a larger water surface area towards the rear of the boiler than the front, because a given height of water leaving the front of the boiler will result in a smaller increase in height at the rear.

Regarding the boiling in the gauge glass, I don't get to play with 60007, but on miniature locos with injector feed to the backhead, I am quite accustomed to seeing the water level in the glass suddenly stop dancing when the injector starts feeding, where before it will have been anything but steady. Maybe 60007 has a certain circulation pattern established when running in reverse that favourably brings hot water to the left side of the backhead (for some strange reason) but when it's been getting hot for a while the behaviour spreads across. Just a thought, but is the fireman's side injector usually used in preference to the driver's side, so water feeds in on one side of the barrel?

 

There is a manifold for auxiliaries at the rear crown of the boiler but I can't remember what is taken off this. The injector steam feeds are from internal pipes so well away from the backhead. I think the ejector is from the manifold but my memory could be quite wrong.
To answer 8126's query about the injector feed, yes, the fireman's side will get the most use but it is far from exclusive in this respect. I think the feed pipe goes well down the barrel but not certain. I've been searching for a boiler drawing which shows these things but with no success.

 

Sorry, the comment about backhead feed was a bit of an irrelevant tangent. I know the A4 feed goes a reasonable distance from there, but I wondered if having the cold feed in preferentially on one side influenced the circulation in the boiler, such that hotter water was to be found on the left side of the backhead, and thus exhibiting more boiling in the gauge glass.

The alternative may be that the LHS gauge glass fitting on 60007 protrudes more or less far through the backhead, compared to the RHS glass and most other engines, for some unknown reason.

 

When I was a chemist, if you had a reaction that clearly occurred but you were at a loss to explain, you'd say "this is caused by a subtle combination of steric and electronic effects" (*) which was a grand way of saying "too complex to understand". Clearly the local conditions of water temperature and pressure in the gauge glass are just tipping you to the boiling side of the equation, but it must be fairly finely balanced given the asymmetry observed.

Obviously there are thermal gradients in the water, and you'd expect the firebox end to be hotter than the overall average; maybe as @8126 says, a slightly different fixing on one gauge glass is just sufficient to put a marginally higher than normal heat input into the water in the glass on the left, so it cools down in the ambient cab temperature fractionally less than expected; maybe in addition there is some take off that also fractionally lowers the pressure in the vicinity of one glass. In other words, I'm tempted to say "this is caused by a subtle combination of temperature and pressure effects".

(*) In chemistry, a steric effect refers to something that is dependent on the physical shape of a molecule; and an electronic effect refers to something that is dependent on the charge distribution. Since that pretty much covers the reasons why chemical reactions occur or don't, saying something was down to a "subtle combination of steric and electronic effects" wasn't far removed from saying "chemistry - it's down to chemistry".

Tom

 

(Continued from another thread).

No, it is two entirely separate effects.

Imagine the loco is stationary on level track; the water shows at a given height in the glass.

Now rotate the loco such that it is at an angle of, say, 1 in 100. Relative to an external observer, the water doesn’t move, but the structure of the loco rotates around the fixed water. To an observer on the footplate, that causes the water level to change in the gauge glass: if the loco is smoke box-first up the gradient, then relatively speaking the water rises in the glass, although in reality the water hasn’t moved, but the back of the loco has instead rotated downwards. That is the effect I am calling the “geometric” effect.

Now move the loco up the gradient at a steady speed. The water is now moving vertically upwards against gravity, so requires a force to be applied to counteract the gravity; that comes from developing a “head”. If the boiler were transparent, the external (ground based) observer would see the head; to the footplate-based observer, it is a second change in water level. This is the effect I am calling the “force based” rise. If the loco is moving forwards up the hill, the result is a double rise of water apparent in the gauge glass. The two effects should be apparent even on a dead loco being towed by another.

Tom

 

I agree about that.

That's where I disagree. Take a glass of water. Tilt it. The water surface remains horizontal so is no longer parallel to the top of the glass, exactly as above. Then accelerate the glass horizontally or along a slope. The acceleration is equivalent to a horizontal gravitational field, so the water surface tilts. Then continue to move the glass but at a constant speed, in any direction: the water surface becomes horizontal, surely?

 

So where is the force moving the water uphill coming from?

 

deleted