Monotube Flash Boiler Design

[TimTaylor]

Troll,

Hot spot tube failures in a water tube boiler are rarely caused by fire side issues - rather they are overwhelmingly the result of water side issue - either scaling that inhibits heat transfer from the metal to the water, allowing the metal under the scale to overheat, or corrosion due to O2 and other chemical components of the boiler water. This is why water treatment, correct blowdown, etc. are important. Yes you can see fire side erosion, but that is usually a design issue related to localized gas velocities and composition (flame impingement).

The goal in any boiler is to have the combustion process as close to stoichiometrically complete as possible before it hits the tubes, then keep it in the heat transfer zone long enough to efficiently extract as much heat as possible. Some excess air is necessary, but you want to minimize it to optimize efficiency - that's why we measure things like stack gas O2, CO and temperature. Optimizing combustion is much easier in a base loaded boiler, where operating conditions and load are fairly constant - on a swing loaded system it can be a lot more challenging.

With gaseous fuels it is relatively straightforward in the design of the burner to assure complete mixing with the combustion air. With liquid fuels, it's a little more complex, and depending on the characteristics of the fuel may require additional measures such as mechanical or steam atomization to get efficient combustion.

 

[Steamchick]

Thanks for the assessment of my comments. I reckon you are better in control than we know, but for sake of trying to help, I just propose ideas that may lead to improvements. Not a criticism of your work, just perhaps things you may have not considered.
I have had burners that were very stable in fresh air, then blew back to burn at the jet due to back pressure. More recently I have read a few papers that give reasonable guides to expending cross-sections through burners and boilers and these explain my experience, good and bad. It is not correct to follow a constant cross sectional area from burner / orifice through to final exit from the chimney/ exhaust. Better to follow the planned model ratios as this improves overall system performance. As you will appreciate when the burner is inside the boiler with energy extracted from the combustion gases, etc. Hope you have it right first time!
K2

 

[Steamchick]

One thing not covered is cavitation erosion... I know nearly nothing about this, but a worked case I was involved with has an indication of something that may have a similarity to fluids boiling in boiler tubes.
The incident occurred with an oil-water heat exchanger.
Hydraulic oil, up to 150deg.C. against a pressurised water glycol mix, that boiled when it should not. The steam bubbles, with high flow through the heat exchanger, caused cavitation erosion. The fluid should not have boiled. The information limiting application of this system to a 50% glycol mix had not been relayed, so the system was filled with only 33% glycol, which permitted the boiling and cavitation erosion. Correct addition of gloom concentration cured the issue.
The correlation in my brain with this boiler is that to avoid chemical breakdown of the boiler fluid, the local temperature must never exceed a value. So does that mean avoiding boiling locally to the hottest parts - near to flames? I.E. A high flow and pressure to prevent boiling by adequately "cooling" the hottest zones?
I have no idea how to model temperature through a boiler, or how to determine the point where liquid becomes a gas...
K2

 

[TimTaylor]

Steamchick,

Your suggestion of an O2 sensor in the boiler exhaust is a good one.

It looks to me like Toymaker is setting up his burner control as an exclusively feed-forward algorithm. While you can do this in theory, in the real world there are just too many variables to compensate for. I think people often forget that combustion is a mass based reaction. Minor changes in things like combustion air temperature and humidity, and fuel temperature and viscosity can have significant effects. Adding an empirical element via s simple feed back measurement makes it a lot easier.

 

[TimTaylor]

One thing not covered is cavitation erosion... I know nearly nothing about this, but a worked case I was involved with has an indication of something that may have a similarity to fluids boiling in boiler tubes.

These are design issues. Sometimes you can compensate via operational control changes, other times you just have to bite the bullet and make the necessary mechanical modifications.

Case in point: I had a client that used the bow thruster on a large vessel for operational steerage control. They wanted more control, so decided during an annual overhaul to replace the 1000kw motor with a 1500kw motor and change the pitch on the impeller to move more water. Unfortunately, no one bothered to check the design limitations of the existing tunnel. Essentially they were trying to put 1500KW of water through a 1000KW tunnel. The resulting vibration at higher run speeds due to cavitation, caused bolts to break and welds to crack. Another issue was that the impeller blade pass frequency coincided with a motor resonant frequency (FTF), further amplifying the vibration amplitude when cavitation occurred. On top of that, the degree of cavitation was directional, as the impeller was offset in the tunnel as it was driven by a right angle gearbox. Our interim solution was to limit the operating speed of the thruster and add a vibration sensor/alarm. The ultimate fix was to replace the entire thruster assembly (new tunnel/gearbox/ impeller) at a subsequent overhaul.

 

[The Ignoble Troll]

Troll,

Hot spot tube failures in a water tube boiler are rarely caused by fire side issues - rather they are overwhelmingly the result of water side issue - either scaling that inhibits heat transfer from the metal to the water, allowing the metal under the scale to overheat, or corrosion due to O2 and other chemical components of the boiler water. This is why water treatment, correct blowdown, etc. are important. Yes you can see fire side erosion, but that is usually a design issue related to localized gas velocities and composition (flame impingement).

The goal in any boiler is to have the combustion process as close to stoichiometrically complete as possible before it hits the tubes, then keep it in the heat transfer zone long enough to efficiently extract as much heat as possible. Some excess air is necessary, but you want to minimize it to optimize efficiency - that's why we measure things like stack gas O2, CO and temperature. Optimizing combustion is much easier in a base loaded boiler, where operating conditions and load are fairly constant - on a swing loaded system it can be a lot more challenging.

With gaseous fuels it is relatively straightforward in the design of the burner to assure complete mixing with the combustion air. With liquid fuels, it's a little more complex, and depending on the characteristics of the fuel may require additional measures such as mechanical or steam atomization to get efficient combustion.

I 100% agree.

 

[The Ignoble Troll]

Steamchick,

Your suggestion of an O2 sensor in the boiler exhaust is a good one.

It looks to me like Toymaker is setting up his burner control as an exclusively feed-forward algorithm. While you can do this in theory, in the real world there are just too many variables to compensate for. I think people often forget that combustion is a mass based reaction. Minor changes in things like combustion air temperature and humidity, and fuel temperature and viscosity can have significant effects. Adding an empirical element via s simple feed back measurement makes it a lot easier.

IIRC a change in two degrees Celsius and an increase to 80% humidity costs a Honda gx196 about 1 hp out of 6.5hp.

 

[HMEL]

First, thanks for the detailed info on combustion processes, very enlightening info concerning the eroding effects of blue flames.

The primary reason I've focused on producing a blue flame is fuel efficiency. The yellow flames seen in my earlier designs quickly coated everything they touched with a thin layer of black soot, which is a clear sign of incomplete combustion; the blue flames leave no residue, no black soot, indicating complete combustion. Since my end goal is to place the final engine in a car, fuel efficiency is very important to me.

Seems I have a choice. I can use blue flames and expect a shorter lifespan of the boiler tubes. Or I can use yellow flames and live with lower fuel efficiency and soot build-up inside the boiler.

A blue flame is a result of complete combustion and it yields the highest temperature. If you do not allow flame impingement on the heat transfer surface there will normally be no problem provide the fuel does not contain sulfur or nitrogen compounds which form acids. You will make more NOX since the temperature is high. Flame impingement is the worry or should be of boiler designers. I have seen issues with just about every type of boiler from improper air flow to poor air fuel ratios. This can occur using goal, fuel oil or natural gas and its equivalents.

 

[The Ignoble Troll]

Out of curiosity, have you been measuring the heat loss through your burner walls as you re-modled it?

A reverse airflow over it, like helicopter engines often use, with the air being drawn into the burner by venturi might net you some fuel efficiency gains if the thing is running really hot. It would also let you potentially burn more fuel with less blower power.

 

[Toymaker]

I would actually respectfully disagree with you, but I'll address that lower down.

On erosion, you hafta look at when you have. You have a forced air burner with very high flow rates.

Take two examples with the same color flame, one is a USA space shuttle main engine, the other is a methanol spirt lamp.

Both flames are similar in color and are blue with high IR signatures.

Both are so faint that they are invisible in daylight.



The shuttle hydrogen/oxygen engine flame is clearly going to erode anything you part in front of it.


The soft methanol flame is going to clearly not erode materials put above it, unless they are very fine like transformer wire but that's extending the metaphor beyond meaning.



You built a jet engine combustion can. You pretty much have an electric Thermojet.



You need to look at the flame as something harsh, turbulent and fast moving, not at all like a stove flame.


On the note of burner flame color; soot, smoke, flaming black snow flakes (that's sorta what droplets of unburnt fuel look like as they eject with the flame) and so on, are the indicators of incomplete combustion.



As I mentioned earlier, I have and it is possible to burn the same fuel air mixture with all those colors - which means they are not indicators of completion of combustion efficiency but flame front speed.



I betcha that given your fuel flow rate that you could make a less efficient burner and get a two foot long yellow soft flame sootless. You've taken that and made it a few cm.


That means that the energy intensity, both thermal, kinetic and chemical are highly focused.

Imagine that you are looking at two 100% efficiency fixed displament pumps. One moves 1714 gallons at 1 psi the other moves 1 gpm at 1714 psi, each take exactly 1 horse power.


Which will cut off your finger?


For you, "combustion efficiency" should mean both that you have complete combustion as measured at the exhaust (easy to determine with a Grove multi gas sensor) AND that you are spreading that flames heat out for the whole boiler to absorb, because a Hotspot could be catastrophic.

Going waaaaaaaay back to your surmise that poping a leak will mean a jet of nasty gas not an explosion; I would personally bet on damage to the refrigerating cause by a hot spot to form an autocataltic reaction that could cause a broad failure as tlarge sections of the boiler corrode internally vs a pin hole. Aka if the seals don't go first, you could have an actual explosion realesing stuff that is more toxic then cyanide.


The web is sadly a midden for "blue" flames.

I'm not trying to preach or put you down or anything. Your project is well beyond my machining and electronic abilities. You are very smart, and it's clear.


I am just giving my .02$ as someone who spent a lot of time working with flame chemistry and is worried you may get injured due to the web's poor info.

All the best.

-Troll

If I understand most of your post correctly, your overarching theme is, "be cognizant of energy density"; which is good advice which I do appreciate.

I do agree with you that my forced air burner has a much higher energy density than a typical gas stove. So, taking the pot & stove analogy a bit further: lets assume the aluminum pot contains some amount of water,...even after the water reaches a rolling boil, the steam will carry away enough heat (BTUs) to prevent the flame from damaging the metal pot. However, once the water is gone, and the pot is dry, even the low density blue flames from the kitchen stove will melt the pot. This example tells me that as long as the metal pot is kept cool enough, that a flame's heat density great enough to easily melt an uncooled pot, will not be able to damage a cooled one.

Lets look at an example using a propane torch. Have you ever tried to braze or even solder a copper water pipe with a little water still remaining in the pipe near the end you're heating? As long as a bit of water remains in the pipe, it will prevent you from heating the pipe to the temperature needed to melt the solder. My point: as long as adequate cooling is in place, it's pretty much impossible to overheat a metal surface.

Final example: A few months back, I tested an aluminum structure I named, Radiant Disk. This disk was designed to be located directly in the path of the flames from the burner where it would transfer large amounts of radiant & convection heat energy into the water (working fluid) inside the disk. It worked great!!! Unfortunately I was never able to prevent leaks from forming between the aluminum disk and the brass fittings going into the disk, so the disk is not used in the boiler today. However, I did test the disk to ensure the flames would not melt the aluminum fins I had machined into the surface. All the tests I ran on the disk were short and the cumulative time was likely under 10 minutes,...but a close examination of the disk, using a magnifying lens to allow me to spot the smallest pits, revealed absolutely no heat damage. Honestly, I expected to find some heat damage around the sharp edges on the fins, but the only changes I could find from the flames was a tiny bit of soot.

Perhaps after many hours of operation, the aluminum fins would show some damage,...perhaps. But these tests have left me feeling confident on my design.

 

[Toymaker]

Out of curiosity, have you been measuring the heat loss through your burner walls as you re-modled it?

A reverse airflow over it, like helicopter engines often use, with the air being drawn into the burner by venturi might net you some fuel efficiency gains if the thing is running really hot. It would also let you potentially burn more fuel with less blower power.

The entire hot section of the burner assembly is located inside the boiler's copper tube coils, (see diagram in post #437). Any heat "loss" through the burner walls results in directly heating the boiler tubes

 

[Toymaker]

Hopefully, and with lots of luck and helpful suggestions from forum members like yourselves, at some future date I will have a functioning steam turbine engine. Once I reach that point, I will go back and install lots of additional sensors, like O2 and blower air pressure, and use those inputs as feedback for the ECU. So keep making those suggestions, as I will likely use them in the future,.... but for now,...I'm focused on just getting a bare-bones steam turbine spinning

 

[Steamchick]

This exciting and challenging project is certainly attracting a lot of expert interest and comment, so I hope you continue to entertain us with your developments. Excellent work! Keep it up!
Thanks,
K2

 

[The Ignoble Troll]

If I understand most of your post correctly, your overarching theme is, "be cognizant of energy density"; which is good advice which I do appreciate.

I do agree with you that my forced air burner has a much higher energy density than a typical gas stove. So, taking the pot & stove analogy a bit further: lets assume the aluminum pot contains some amount of water,...even after the water reaches a rolling boil, the steam will carry away enough heat (BTUs) to prevent the flame from damaging the metal pot. However, once the water is gone, and the pot is dry, even the low density blue flames from the kitchen stove will melt the pot. This example tells me that as long as the metal pot is kept cool enough, that a flame's heat density great enough to easily melt an uncooled pot, will not be able to damage a cooled one.

Lets look at an example using a propane torch. Have you ever tried to braze or even solder a copper water pipe with a little water still remaining in the pipe near the end you're heating? As long as a bit of water remains in the pipe, it will prevent you from heating the pipe to the temperature needed to melt the solder. My point: as long as adequate cooling is in place, it's pretty much impossible to overheat a metal surface.

Final example: A few months back, I tested an aluminum structure I named, Radiant Disk. This disk was designed to be located directly in the path of the flames from the burner where it would transfer large amounts of radiant & convection heat energy into the water (working fluid) inside the disk. It worked great!!! Unfortunately I was never able to prevent leaks from forming between the aluminum disk and the brass fittings going into the disk, so the disk is not used in the boiler today. However, I did test the disk to ensure the flames would not melt the aluminum fins I had machined into the surface. All the tests I ran on the disk were short and the cumulative time was likely under 10 minutes,...but a close examination of the disk, using a magnifying lens to allow me to spot the smallest pits, revealed absolutely no heat damage. Honestly, I expected to find some heat damage around the sharp edges on the fins, but the only changes I could find from the flames was a tiny bit of soot.

Perhaps after many hours of operation, the aluminum fins would show some damage,...perhaps. But these tests have left me feeling confident on my design.

View attachment 155149

If I want to join a water filled pipe I open one end, whip out my oxyfuel system and hit it with copper phosphorus rods . Water be damned.


As mentioned by other posters, it's direct flame impingement on your boiler tubes that worries me...

It all boils down to the low decomposition point of the working fluid.


Just out of curiosity, have you considered a mono chemical non-halogenated fluid like R-290?

Or any of the blended hydrocarbons?


Once upon a magical time, there were engines that burned basically heptane but also used it as the pressurized working fluid in the boiler/piston engine.

https://en.m.wikipedia.org/wiki/Naphtha_launch

 

[Toymaker]

If I want to join a water filled pipe I open one end, whip out my oxyfuel system and hit it with copper phosphorus rods . Water be damned.


As mentioned by other posters, it's direct flame impingement on your boiler tubes that worries me...

It all boils down to the low decomposition point of the working fluid.

There's a short expression used among steam automobile enthusiasts I rather like; "Keep the Fluid moving!!" It means that as long as the working fluid is moving through the tubes fast enough to carry away the burner's heat, the boiler tubes will be fine. I'm hoping that keeping fluid velocity on the fast side will also help prevent decomposition, but honestly, that's TBD.

Just out of curiosity, have you considered a mono chemical non-halogenated fluid like R-290?

R-290 = Propane;....highly flammable fluids are most definitely NOT on my list of possible working fluids. Also, I prefer a liquid with a boiling point above room temperature.

Or any of the blended hydrocarbons?

Yes, both XP30 (aka R-514a) and Opteon MZ (aka HFO-1336mzz-Z) are possible choices. However, I don't know if I can buy them here in Thailand.

Once upon a magical time, there were engines that burned basically heptane but also used it as the pressurized working fluid in the boiler/piston engine.

https://en.m.wikipedia.org/wiki/Naphtha_launch

 

[Steamchick]

I understand the "keep the fluid moving" motto, just don't understand the maths to predict the stage at which the fluid boils and changes form Liquid to Gaseous. I'm sure there is a mathematical Model on the web somewhere... so you can plug-in your sizes and dimensions and fluid properties, heat input etc. and get a projection of what may be happening?
Most interesting!
K2

 

[Steamchick]

Troll,
I am curious to know about: "my oxyfuel system ..." - I mean "yours..".
I clart-on fixing old copper boilers using a pair (or 3 or 4) petrol blowlamps, as they are the convenient high intensity and high power flames I have for the price of a pint of petrol (£1.42/litre). = MUCH less than Propane (£6/litre!). ANd with multiple blowlamps I can direct heat for pre-warming the boiler below silver soldering melting point and then use on just on the point to be repaired, without having a single HUGE burner/flame cook everything.
I store the petrol in my Moto Guzzi fuel tank.... where it is mostly used.... Very convenient!
K2

 

[Toymaker]

I understand the "keep the fluid moving" motto, just don't understand the maths to predict the stage at which the fluid boils and changes form Liquid to Gaseous. I'm sure there is a mathematical Model on the web somewhere... so you can plug-in your sizes and dimensions and fluid properties, heat input etc. and get a projection of what may be happening?
Most interesting!
K2

You're not alone in not fully understanding the math needed to predict where a given working fluid will change state from liquid to vapor; I know I don't, and I suspect there are only a few people in the industry that actually do fully understand the math.

I believe what's needed is a software package capable of performing CFD (Computational Fluid Dynamics); CAD packages such as AutoDesk and Solid Works both advertise to have these capabilities,...but these are high end software utilities that I don't have.

Google searches tell me there are several free CFD utilities available online, and most reviews say OpenFOAM is the best. Perhaps one day I'll take the time to learn how to use it

 

[Steamchick]

Just tried "clicking the buttons" - But reading the instructions (I have a Windows computer) it seems I need to figure out what software I need to allow Windows to rum Linux programs, download a few packages, and convert each package with some stuff to make the whole thing work.
A bit beyond ANYTHING I have attempted before...
So I'll give this a PASS.... and go back to my simple sums...
Sorry Folk! I just don't feel the need to study CFD and LINUX computing from scratch today! (Failed computing in my degree in 1975... when computers were rooms of people with punch cards! and cabinets with tape machines inside! - Never looked back and went there again. Now it's "on the phone" - I think?).
K2

 

[The Ignoble Troll]

Troll,
I am curious to know about: "my oxyfuel system ..." - I mean "yours..".
I clart-on fixing old copper boilers using a pair (or 3 or 4) petrol blowlamps, as they are the convenient high intensity and high power flames I have for the price of a pint of petrol (£1.42/litre). = MUCH less than Propane (£6/litre!). ANd with multiple blowlamps I can direct heat for pre-warming the boiler below silver soldering melting point and then use on just on the point to be repaired, without having a single HUGE burner/flame cook everything.
I store the petrol in my Moto Guzzi fuel tank.... where it is mostly used.... Very convenient!
K2

I have a coveted supply of MPS, that plus an oxygen tank is going to get just about anything small, like a copper pipe, hot enough to braze.

I miss MPS like a lost lover. Sigh.