Monotube Boiler

[vtsteam]

Well, after probably a decade of thinking about it, and one unsatisfactory attempt at building a wood fuel monotube boiler 5 years ago, I'm going to start again. I almost hesitate to bring up monotube boiler construction here since there is a lot of criticism of monotubes in general to be found on the internet by people of wide experience in the steam engine field. And their objections are quite helpful and instructive.

Nevertheless I have to have a go at it. My primary interest in the type is due to safety concerns: the commonly agreed difference in the released energy of a monotube rupture in a system containing only a small amount of water compared to a conventional boiler vessel rupture with a large superheated water content and huge potential energy reserve.

A second reason is the simpler fabrication requirements.

And the third is the desire to use wood and solid fuel to produce electricity. It's simple and possibly selfish reason. I have a lot of wood. I don't have an oil field or refinery, or a coal mine. And possibly less selfish reasons. I think renewable fuels are a human species survival advantage over the continued  use of fossil fuels. Even if someone disagrees, I'm sure they will applaud my use of alternatives -- leaves more petroleum for them to use!! Win/win, right?

Finally, fourth reason -- I like a technical challenge.

I always believe in starting with basics -- and that means the basic assumptions about what I want to do. I think that very often the criticisms leveled at these kinds of projects are general in nature or attached to experiences with very different starting assumptions. Thus criticisms of oranges are leveled at apples -- or fruit in general. So it's important to understand in advance what the assumptions and purposes are.

As best I can express them or think about them, they are for this very special circumstance:

1.) a monotube boiler which works well with a wood heat source. I have in mind a wood burner I have developed over the years which has a relatively stable heat output and is throttleable

2.) capable of delivering a maximum of 5 shaft horsepower. I hope from a converted 5 hp 4 cycle engine I am also building. (This does not mean the engine WILL be capable of that -- just that the maximum steam capacity of the boiler should be capable of that level of steam production -- in other words this is a boiler sizing specification, not a performance prediction).

3.) that the steam temperature be optimized for 400F (200C). If this means that the engine I'm working on can not possibly attain 5 hp, then fine, this temperature is still a max design running temp. I do realize that flash steam generators can and will generate much higher temperatures and pressures briefly (or permanently!), and that they are difficult to control and stabilize. I just mean that this is the max design running temperature for steam output.

4.) that the electrical generator attached to the steam engine is a DC generator for charging a battery. This means a relatively stable steam output requirement. This is not a monotube boiler intended for a vehicle, or load following AC output requiring rapid throttle response, or highly regulated RPM for frequency stability..

5.) that the system is designed for occasional or emergency use, not 24/7 off grid electrical generation.

6.) that this be fun and interesting, educational, and reasonably safe as an experimental design process, and that if any or all of the goals and specifications are not reached, I will at least have the experience of trying and discovering interesting things along the way, even if they have been already discovered a hundred times before.

So that's what I'm about with this. I'm just a mad modder.

[doubleboost]

I shall be watching with great interest
John

[vtsteam]

Thanks John  .

Okay so I'm going to just type out a bunch of thoughts and observations and readings, all mixed up, because to even start on something like this it's important to just throw a bunch of junk out on the table and see what you've got -- like rooting through the scrap box of observations.

The first thing that hits me is that there are three major geometric orientations for the tubes I've seen.

The first is found in model boats, including highly developed racing steam hydroplanes. They have horizontally oriented spirals.

The implication of this is that they do not and can not maintain a "water level" since the coils are circular in the vertical plane. If the pumping speed is high enough, we can assume a fair number of the early heated sections of the coil may nevertheless be filled with water, and later sections steam, but that isn't literally a  water level.

It should be added that one modification of this form by HH Groves (and in fact one of the earliest -- dating back a century) has in addition a series of lengthwise runs.

The second major type of monotube is a vertically oriented spiral often found in steam cars. In these types people frequently speak of and attempt to maintain through various sensors and controls a literal "water level"

I don't know whether it is the compactness enforced by their use in an auto, or the influence of traditional boiler types that this vertical axis form and literal water level concept is adopted. Or perhaps the higher horsepower needs of an automobile, compared to a model boat, means that this boiler style is most applicable in larger sizes. Be that as it may, the needs of compactness, horsepower, etc do not seem to apply to my own set of requirements outlined above. So I'm open to thinking about the other, horizontal form as well as this one.

I read that many of the difficulties described by monotube experimenters center around failure to maintain water level, and the varying point of transition in the monotube to steam is frequently discussed as a cause of these failures. So that is a focal point to think about.

The third form of monotube is just a random bending of tube -- basically a purpose made rats nest of tubing, more or less rising vertically in a casing. I saw online one very small boat that had this arrangement, and the builder said he thought the chaotic arrangement was favorable to heat exchange, and worked well for him.

I guess I'm perverse enough to like that kind of thing! It appeals to me, though I don't know if I'll do things that way. But I have to admit, I'm attracted to it because in its odd way, it makes sense.

I do wonder what it does to the concept of a water level, and whether the "coils" rise progressively through the casing, or weave up and down randomly, as the horizontal coils of a model boat do in more regular fashion. Don't know the answer to that yet.

Finally there is the question of flow and counterflow for the water and steam in relation to the flow of hot gasses from the burner. Counterflow seems the most favored, though means that the tubing is subjected to greatest heat where it isn't cooled by entry water. To me this seems like a recipe for a relatively unstable system by comparison with a parallel flow system, though the latter will not be able to pick up as high a maximum temperature as the former. But if we are limiting temperature anyway, Maybe stability is more important.

I don't know. Something to think about.

[vtsteam]

So now another thought strikes me -- a parallel flow system puts the coldest water in the hottest gas, and so must effect the quickest heat exchange, initially. That exchange rate slows rapidly as the water heats into steam and the gasses cool rapidly having given up that heat. This should happen relatively early in the tube length.

If the temperature output requirement is relatively modest, then it seems to me that it can be achieved in a shorter length of monotube with parallel flow than counter flow. So the possible considerations are:

1.)Lower maximum temperature
2.)Shorter monotube
3.)Higher natural system stability
4.)Better tube protection.

[vtsteam]

More thoughts.

The burning out of tubes is not a simple result of exceeding the melting temperature or exceeding the yield strength at elevated temperature of the tubes.

It can also be the result of literal burning, oxidation at high enough temperature. This was brought home to me in my early experience of metal casting in a charcoal furnace. I had used a stainless steel container as a crucible for aluminum maybe 5 times before it developed a leak at the bottom corner.

What gives, I thought? It's stainless steel -- it shouldn't melt!

Well there was an excess of oxygen near the tuyere where the blower entry was, and early on, I used to turn the blower up high to get as quick a melt as possible. What was actually happening was that I was literally burning the stainless with excess air. The same principle as the oxy-acetylene torch when cutting steel. It is possible, after a cut is started, to completely turn off the flow of acetylene, and cut with pure oxygen alone. The iron itself is the fuel used to continue the melt.

Another example is heating a copper pipe with an oxyacetylene torch when brazing. An oxidizing flame will leave the copper covered with rough scale. A reducing flame will actually remove scale and corrosion and convert it into shiny copper.

So, it seems to me that in the interest of long tube life, I would like to put the tubes in a section of the boiler where the combustion is complete, and has used up all of the oxygen drawn into the burner. Personally, I don't like putting any kind of heat exchangers in combustion chambers in general. They cool the combustion, which is rarely an advantage. I like to site them in the exhaust.

This might also make the use of copper tubing feasible -- it is often objected to on the grounds that it scales badly in the flame of the burner.

So the following considerations might apply to our new monotube boiler, unless some better ideas come up:

1.) avoid excess air
2.) locate tubes out of the combustion zone if possible, or where in the zone, make sure it is water cooled. Again this favors a parallel flow approach.
3.) copper may be feasible if oxidation can be controlled.

[John Hill]

What scope is there to operate the boiler at pressure such that the contents of the boiler tube are liquid at all time and only flash to steam as they leave it?

[vtsteam]

Hi John!

In this  particular project that isn't a goal, though I don't see why it couldn't be for another person's work. In fact the superheated water can be maintained all the way into the cylinder chamber with the use of an injector. See: http://www.flashsteam.com/

Any project of that sort would be extremely interesting and admirable to me!

My own goal is to fulfill the original set of specs and assumptions in the first post in as basic and mechanically simple a way as I can work out, with the most easily available materials I can find, at the lowest cost I can.

That brings up the topic of efficiency in a roundabout way. Efficiencies of steam production and usage form the basis of many interesting debates on the Internet.

In general the efficiency I strive for is not the ultimate thermodynamic efficiency of steam generation and power production, but the efficiency of cost and physical means per watt within the constraints of the project specs and assumptions That guides practically every choice I make.

As an example, the choice to use wood as fuel is a choice based on the fact that it is available and free for me, not on whether it is the ideal fuel for producing steam with the most efficient burner. Likewise other choices which will eventually come up.

[vtsteam]

More thoughts on flow direction.

Thinking about horizontal vs vertical coils, and the randomly oriented rat's nest style tubing in the SL Alba. (remembered the boat name, but not yet the owner apologies to him  )  it seems to me that there are two very different mechanisms here beyond just parallel and counter flow. designs

And those are gradient and non-gradient hot gas flow in relation to the tubing.

Both parallel flow and counter flow depend on an assumed gradient of heat parallel with the overall tube direction. But what happens in a randomly oriented monotube?

There is no gradient overall, though locally there will be hotter and cooler locations along the tube. Similarly in HH Groves horizontal style monotube boiler with its axial weave back and forth within the outer coils, there is no overall gradient.

This lack of gradient may be an assistance to to stability.   .......thinking about this......

[Brass_Machine]

Well... I don't have much experience with a boiler. I have only a tiny one that came with my first steam engine kit that I built. So I will not be able offer much assistance. I do have a piece of 3" heavy walled copper tube that will be used for one at a later date.

I will be watching with great interest.

Eric

[vtsteam]

In a vertical oriented automotive type monotube, one of the basic problems I've seen discussed is that when an increase in steam is suddenly called for, extra cold water is drawn into the bottom end by the reduction in pressure at the the steam chest end. This addition moves the steam transition area further up the tube, reducing the  steam generation and pressure, compounding the need for more steam. If the throttle is opened further, pressure reduces further. Unless more heat is added. Well, of course it is in most practical systems, but the lag in response as well as overheating at the top end (closest to the burner can cause further problems, including tube failure.

This seems like a classical problem of initial negative stability in the monotube boiler system. A displacement from equilibrium  increasingly reduces the tendency to equilibrium. And it seems to me that a conventional counter flow gradient monotube will tend to accentuate this negative stability. The solutions require sophisticated control mechanisms which enforce stability, and avoid damage. These are difficult to produce.

On the other hand, a randomly arranged tube would seem to reduce the unstable tendency -- or at least increase the rate of recovery (if the throttle is not opened further). In a random flow monotube, increasing the heat available would not primarily affect the "top" end (there is no top end) but would be applied throughout the monotube in a relatively even fashion.

The increased load of water created by low pressure will have a higher heat transfer rate across the tubes since more of the tube is cooler in temperature, and the water is a much better conductor than steam, Because transfer is faster, Reduction in the flue gas temperature will occur more quickly, and response to increased burner output will therefore be more rapid. In other words, response lag will be reduced.

Now what about parallel flow?

A sudden increase in water volume in the tube, would seem to me respond even more quickly to an increase in burner temperature. And the cooling effect of the added water in the tubes (now closest to the burner) would tend to lower the temperature of the steam chest end of the tube, rather than overheat it, thus more more effectively avoiding damage.

These are all theories. I really don't know the answer. And It's pretty likely to have been discussed by others before. So apologies to those who may think this is old hat.

Specifics of an individual boiler will be more important than theory in taking advantage of any theory like those thoughts above -- if it is an actual advantage at all.

[vtsteam]

Brass Machine thanks!!
I think I'd favor a conventional boiler for stationary models, too. I have a few I'd like to get working some day.

I hope this thread will at least be interesting, even if the results don't pan out.

[vtsteam]

So to conclude -- the big problems I've been thinking about in monotube boilers are tube damage from overheating, tube damage from excess air, control and stability problems in steam production, cost inefficiency, exotic materials requirements, and system complexity.

What I'm tending to favor to try to mediate the problems are:

1.) a horizontal design
2.) a parallel flow or random flow design
3.) a heat exchanger located largely past the combustion area
4.) reduction of excess air
5.) wood fuel
6.) a relatively constant load style of electric generation

These are not the common choices in monotube construction or design today, in fact added together they are pretty much the opposite of conventional wisdom. It is not the way to do things if the greatest value of heat exchange and temperature is desired from a monotube. However, it might actually be a practical response to the conditions and assumptions I gave in the first post, and so efficient in that sense. In other words, it might be good enough for who it is for.

[R.G.Y.]

Beeing a boat man I think the horizontal design will have a  lower CofG, for a longer coil as in steam hydroplan racers.

[vtsteam]

Yes RGY, that's very true. 

And it can also be longer if it is horizontal and still fit tween decks. Which can mean easier heat conversion over a longer length of heat exchanger.

On a slow boat, though, we need a stack, so lose some of that layout advantage (unless there's a blower).

Fast boats like racers use ram air to provide oxygen to the burner, flow through the exchanger, and to clear exhaust.

[fcheslop]

http://www.modelboatmayhem.co.uk/forum/index.php?topic=24568.0
http://www.modelboatmayhem.co.uk/forum/index.php?topic=15817.0
There maybe some info  in these threads and E T Westburys Experimental Flash Steam book is pretty good
I did play with flash steam in the early 80s before the advent of cheap electronics and found for marine use a simple figure of 8 worked well.

Good luck
best wishes
frazer

[vtsteam]

Frazer, wow thank you  great information in those two threads! It's going to take me all night to read and absorb.

I've had ET Westbury's flash steam book and Benson and Rayman's experimental flash steam book for almost ten years  --  they are well worn! I have also collected Model Engineers for many years and have paid close attention to Dr Chaddock's and Jay Bamford's work and Bob Kirtley's among others. R.F. Yates also collected a lot of monotube information in his book "Model Making." And even Henry Greenlee had a go at it with a flash domestic power plant. And then of course HH Groves, and his steam aircraft, who seems to have hit many advanced concepts all at once in his engines and very early on. Kind of like Parsons did in turbines. It's like these guys put 6 different major innovations in any particular step they make, and they just work out of the box.

Anyway the above two threads are gold.

Thank you again! 

[dvbydt]

Another web site I have bookmarked from a few years ago :-

http://www.flysteam.co.uk/index.htm

Might still be of use, he mentions the same names as you did.
I want to make a small version (eventually), to run my models.

Ian

[doubleboost]

If
You lads dont stop posting links i will be up all nite reading
John

[fcheslop]

I don't know exactly what the boiler is for is it for a model or for a small launch?
The use of PICs has had a great effect on boiler control I keep thinking of getting back into it but would consider some kind of feed back loop.Don't know what you think ?
I seem to remember a couple of American designs in The Modern Steam Launch magazine I think the editor was Bill Fitt or Durham
Also I think Lunevalley in the UK made a monotube boiler there maybe some info on the Steam Boat Association UK site
http://www.steamboatassociation.org.uk/Default.aspx?pageId=1242242
Bobs drawings for Pisces 2 are available from My Hobby Store
cheers

[vtsteam]

A generator, Frazer, the specs are in the first post.

I think electronic control using PICs is fine and interesting, very attractive and I really look forward to seeing more. But for me the need for a number of sensors and actuators and a microprocessor is a different direction than the one I want to head in. Control by wire is I think  very applicable to a system with negative stability, where highest heat exchange and high temperature performance is the goal.

I would love to see your work if you pick back up on that direction. That would be cool!

I'm taking the opposite tack which is dumbing down the system and reducing the heat exchange efficiency to try to get more control stability. Most of what I am thinking about has no application to racing or automotive steam power. And so it is kind of interesting in its own right.

As far as I know, the only surviving common practical domestic use for steam power are the pressure cooker and espresso maker. Both of those use simple mechanical control rather than electronic control of the steam itself -- and both are robust systems.

I'm a sentimentalist. I still don't have DRO's on my tools, though I'm sure I will some day. Just not ready for it yet -- having fun acting like I'm back in the last century still!

[vtsteam]

Another web site I have bookmarked from a few years ago :-

http://www.flysteam.co.uk/index.htm

Might still be of use, he mentions the same names as you did.
I want to make a small version (eventually), to run my models.

Ian

dvbyvdt thanks. I have seen that site, and it's very good. In fact it had an earlier incarnation a number of years ago, and I wrote Geoff back then asking a bunch of questions about monotubes. He said he'd been working on a mechanical control mechanism and a book on monotubes. I think he even sent me a draft chapter. Besides all of the interesting stuff he has now on the website (it is much larger now than it was) one of the most important to me is the discussion of tubing diameters and the relationship of water velocity inside the monotube.

There seemed to be a magic number for diameter in both his readings of tests by Edgar Westbury in Model Engineer and his experience with a small steam launch. And that was 3/16" diameter tube. His theory was that the velocity of the water in the tube (for the sizes of engine in the small range) was critical to good performance of the monotube.

As a result of this I bought back then a roll of 3/16" stainless steel tube, which I still have. I'm very curious about this point. And we'll probably find out here whether it holds for an engine of increased displacement, like the one I'm adapting. It seems to me that 3/16" may not be a magic number -- but the velocity theory may very well be true.

Seems reasonable to me that the steam needs of a small engine might enforce an optimal velocity with 3/16" tube, and that the steam needs of a larger engine might enforce the same velocity in a larger tube. The reason that 3/16 tube might work well for a pretty good range of small engines is because the next common (imperial measure) increment in tubing size is 1/4", and while that seems only a 33% increase in nominal size, it's a cross-sectional area increase of 77%.

To achieve the same velocity the pumping capacity would need to increase that amount. So the steam volume (and possibly the engine displacement volume) would increase 77% to match.

The tube surface area has only increased 33% however, unless we lengthen it -- so that also might need to increase 33% to bring the heating surface in line with the quantity of water being pumped per minute.

So, basically, going up only one size in tubing has made a big difference in what size engine it would be appropriate for. And so it might seem that for a fairly wide range of small engines 1/4" is too large. Beyond some point, however, I bet that a next size tubing increment is appropriate. And that would logically continue up through the engine sizes.

[John Hill]

Thanks for the interesting read and all those links.  I have incorporated some of what I have learned from them in my latest daft idea which is a "semi-internal steam generator" engine.

[Raggle]

I hope the following is of interest, I do tend to ramble ...

About 40 years ago I made my living as a chemical descaling operator, largely on conventional steam boilers of multiple firetube type. Other work included water jackets on large compressors, cooling towers, etc., in fact anywhere a heat exchange had taken place and deposited calcium carbonate.

I'd already developed an interest in the Doble steam car and I like to think my mention of it at interview helped me to get the job.

One day I was called to see to a Stone's steam (or vapour) generator installed at a knitwear factory in Leicester. I believe it was to provide steam for one or more Hoffman presses, though it may have had other jobs.

I arrived with my full kit of pumps, 70 gallon plastic bath and hoses, and 2 or 3 carbuoys of hydrochloric acid. On being shown the item I confessed to the owner that I had never seen one before and asked if he had a manual. This he presented and left me to it.

I turned straight to the descaling section, piped it up as per instructions and got going. Most of the time is spent waiting between measuring acid strength, adding more as and when required. This gave me ample time to read the rest of the manual.

The Stone's and the Doble are very similar, possibly this later one contained some Doble patents. The fire, kerosene in the case of the Doble, maybe diesel in the Stone's, came in at the top and is electrically blown. The key to steam production is continuous pumping through the coil, which I think was 3 concentric diameters. At a high point above the coils is a steam drum/water reservoir. Steam bubbles are centrifugally seperated in this vessel, inside of which there are 2 water level sensors, high and low.

The pumps are interesting in being of diaphragm type. The diaphragm is at the end of a long pipe and its valves are at the other where the main flow occurs. Thus the diaphragm is not subjected to the full working temperature, being at some distance and merely generating pulses. The pumps are cam driven and the diaphragm can be held stationary by a solenoid. The two pumps looked identical but one is for continuous circulation, the other is the feed pump, triggered by the demands of the 2 sensors in the upper vessel (which I now recall was called the accumulator).

Feed water was preheated by a coil, probably by-pass or parallel connection to the main coil, (or maybe somewhere late down the flue, I can't remember)

I believe that for some models there is a superheater coil somewhere in the steam side, but details escape me.

When acid strength ceased to fall my job was done, other than adding lime to my acid bath to neutralise it before dumping down the drain and a number of clean water flushes and I packed up my truck without having to do the usual cleanup of my work area.

The owner reassembled the burner system and pressed the start button. 100 psi arrived in a few minutes which he pronounced good.

The unit by the way was about 3 ft high and about 3 or 4 ft diameter, not including ancillaries.

Ray

[vtsteam]

That is very interesting stuff, Raggle. Thank you very much for posting it. 

Completely new to me at least are the cam driven diaphragm pumps, the method of de-scaling with hydrochloric acid and watching the pH drop. Some of the other details are more widely reported but the pumps are especially interesting. In fact I'm going WOW, that is just brilliant! Almost every feed pump I've seen in, well, mostly model engineering sources for monotube steam are plunger type. Well D.H. Chaddock experimented briefly with an oscillating cylinder type -- like an oscillating steam engine in reverse.

But thank you so much! I am definitely going to think about this more.

 

[vtsteam]

Reading a second time through I'm struck by the circulation pump and accumulator, as well. And the longevity of this monotube steam generator, its industrial setting, its size. and the 100 lbs working pressure.

Thank you again for this very rare and useful information!