Monotube Flash Boiler Design

[Toymaker]

Actually, under normal circumstances, with the increase of excess air the temperature of boiler exhaust gas increases. Reducing the excess air to the minimum achievable to ensure complete combustion will reduce the exhaust gas temperature. This of course assumes that the boiler has been properly designed for the expected load conditions - over-firing a boiler will always result in increased stack gas temperatures. There always needs to be some excess air to insure complete combustion - how much typically depends on the type of fuel.

Another factor people often fail to consider is that combustion is a mass based reaction - it takes x lbs of O2 with y lbs of fuel to achieve complete combustion. Air density (and the mass per unit volume of it's gas components) changes significantly with temperature, so using cfm only gives you an approximation. Commercial/industrial boilers will use an O2 trim system to compensate for these variations. In situ O2 measurement is the most cost effective approach for small to medium sized boilers. Larger boilers will often add CO measurement to monitor unburned products of combustion to allow optimum control of the process.

Interesting that a small amount of excess air flow would cause higher exhaust temperatures. That's good news for me as the combustion chamber on my burner is air cooled, and cooler metal will lead to a longer life of the burner.

One of the benefits of using DC electric motors to spin a centrifugal air compressor (the leaf blower) is that the motor's rpm is mass flow dependent. When given a fixed voltage and current, the rpm of the small DC motor on the leaf blower, (used to supply air flow into the burner) will vary depending on the load; for an air blower, that load is the air's mass. Low density air on a hot day will allow the motor to spin faster and move a larger volume of air, while high density air on a cold day will cause the motor's rpm to slow down, moving a smaller volume of air. The work being done by the motor is dependent on the mass of the air, not the volume, so a DC motor will move the same mass of air regardless of the air's temperature related density.

The AC motors used on large industrial blowers are typically designed to run at a fixed rpm, regardless of the load, resulting in the higher mass flow on cold days vs hot days, that you've pointed out.

 

[Toymaker]

If you find yourself needing to restrict combustion air, the most energy efficient method (depending on blower design) is usually to damper the combustion air blower inlet, not the outlet.

Water entrainment (wet stream) can be minimized a couple of ways. Most boilers are designed with a steam drum of sufficient volume to reduce the steam velocity, letting the heavier droplets drop out. This is often combined with other design elements such as deflector plates (de-mister section). This can be done internally in the drum, or externally in combination with an appropriately sized steam trap. Another method is to add some superheat to the steam by running the steam from the drum through a coil in the boiler exhaust. Another option is to run the boiler at higher than process pressure and use a pressure reducing control valve, which also will leave a bit of superheat(if there is any there originally) to the lower pressure process pressure steam. In any case using superheat, you need to make sure the pipes are properly insulated to maintain it.

Simply adding a discharge valve to the boiler outlet will not change the moisture content of the steam - the best you can get is D&S. That said, if you are simply running the outlet open to atmosphere now, it will make a difference.

Finally, to get the most efficiency from a turbine, as much of the pressure drop as possible must take place across the blades. You want to maintain the pressure in the steam chest feeding the nozzle(s) as high as possible. Taking a significant pressure drop across the governor valve will significantly reduce the turbine efficiency, as will having too many and/or too large of nozzles.

TimTaylor, an important detail I haven't mentioned recently is that the boiler I'm building will be installed on a mobile vehicle, therefore making everything small and lightweight is an important goal. Dimensions of the burner's outer shell are 10" long x 5" diameter, while the combustion chamber inside the outer shell is only 6" long x 3.25" diameter.
The below photos show the exhaust gases with and without a stainless steel restrictor plate. Fuel and airflow were nearly identical for both pics. There appears to be a fair amount of unburned fuel in the long flame shooting out of the burner as revealed by the orange color. My working theory is that adding the restrictor plate keeps the combustion gases inside the chamber a few milliseconds longer, resulting in more complete combustion.
I welcome your opinion/analysis.

I've considered omitting a governor valve and instead controlling turbine power by controlling boiler output, but I'm concerned that overall system response time will be too slow; this engine will be powering a vehicle and quick response time is an important goal. For now, my plan is to install a control valve and once the entire engine is operational, I can leave the valve full open and attempt to control turbine power output via boiler output.

 

[HMEL]

It's actually not that tough - it's been done for many years in industrial boiler systems. How complex the controls have to be depends on the characteristics of the system. For base loaded systems it is relatively straightforward for both the basic control strategy, and to add feedback trim based on residual O2 in the stack gas. For swing loaded systems, it can be quite a bit more challenging, and you wind up with a stacked control loop configuration with multiple feed-forward/feedback elements. You have to take into account all kinds of variables, including dynamic system response to load changes. It can be really challenging with analog control loops - the advent of DCS systems has made it significantly easier.

A DCS system has a few vulnerabilities that is not often talked, about. One is loss of power, two is shutdown positions on system failure, three is software issues, four is backup transfer, I could go on but a DCS (distributed control system) is nice because it can be incorporated in controlling a lot more equipment. I have experienced most of the first four plus one in which the insurance company made it necessary to hard wire the flame safety system the old fashion way. But I dont think you will see many analog systems being built except for control of emergency shutdown of turbines and electrical gear. It is too costly. The major problem was the people who really did not understand PID, PD, and other control logic application. I once found three filters in a row in the software. But with the right training, the right hardware, the right people it can be a very nice system.

 

[Toymaker]

A DCS system has a few vulnerabilities that is not often talked, about. One is loss of power, two is shutdown positions on system failure, three is software issues, four is backup transfer, I could go on but a DCS (distributed control system) is nice because it can be incorporated in controlling a lot more equipment. I have experienced most of the first four plus one in which the insurance company made it necessary to hard wire the flame safety system the old fashion way. But I dont think you will see many analog systems being built except for control of emergency shutdown of turbines and electrical gear. It is too costly. The major problem was the people who really did not understand PID, PD, and other control logic application. I once found three filters in a row in the software. But with the right training, the right hardware, the right people it can be a very nice system.

To avoid any confusion, my ECU should not be considered a DCS. The computer inside the ECU is an Arduino Mega 2560 microcontroller which is programed using Arduino's version of C++. There are no PIDs, PDs, or PLCs used anywhere.
Excluding software issues, if the ECU were to fail for any reason, all the electric motors instantly stop running. There would be no fuel flow and no airflow into the burner, resulting in no burner flames. Monotube boilers have precious little residual heat in the tubing, so the boiler stops making steam, almost instantly.
Finally, before every start-up, the ECU insures all controlled valves are set to their "start" position.

 

[TimTaylor]

A DCS system has a few vulnerabilities that is not often talked, about. One is loss of power, two is shutdown positions on system failure, three is software issues, four is backup transfer, I could go on but a DCS (distributed control system) is nice because it can be incorporated in controlling a lot more equipment. I have experienced most of the first four plus one in which the insurance company made it necessary to hard wire the flame safety system the old fashion way. But I dont think you will see many analog systems being built except for control of emergency shutdown of turbines and electrical gear. It is too costly. The major problem was the people who really did not understand PID, PD, and other control logic application. I once found three filters in a row in the software. But with the right training, the right hardware, the right people it can be a very nice system.

I never believed in letting the DCS, or analog control systems for that matter, handle flame safeguard functions. We would typically use a dedicated Fireye system that interfaced with the boiler combustion control system. Once all the basic safety interlocks, pre-purge timers, main fuel valve low fire position, etc. are satisfied, the Fireye system handled pilot ignition, confirmed the flame, then released power to the main fuel valve allowing the operator to open the fuel cutoff valve (Maxon or similar), confirm ignition of the main flame, then release modulation control to the combustion control system.

Power failure isn't a big issue with oil/gas fired systems as the fuel safety cutoff valves will do exactly that on power failure, instantly stopping combustion. Even if the fans/pumps stop running it's not really going to create a dangerous situation. Dampers will fail open or to their last position, and control valves will fail to a safe position - open or closed depending on the application. Boilers fired on solid fuel such as wood or coal are a different matter, because when the stoker stops they still have a fuel bed that is burning and releasing heat. These type of boilers are required to have a steam turbine driven feed pump so that as long as there is steam pressure in the drum, there is feedwater going into the boiler.

DCS systems in any critical application are required to have enough battery backup capability to allow an orderly shutdown in the case of loss of mains power. They also have hardware watchdog functions that will detect any software or hardware fault condition and revert control to local manual control stations with bumpless transfer, again allowing a safe and orderly shutdown.

I think the dominate technology for most small to medium applications these days is the PLC. Their capabilities have grown far beyond their original purpose to replace ladder logic relay controls. Even the low end ones have PID and network interface capabilities.

 

[TimTaylor]

To avoid any confusion, my ECU should not be considered a DCS. The computer inside the ECU is an Arduino Mega 2560 microcontroller which is programed using Arduino's version of C++. There are no PIDs, PDs, or PLCs used anywhere.
Excluding software issues, if the ECU were to fail for any reason, all the electric motors instantly stop running. There would be no fuel flow and no airflow into the burner, resulting in no burner flames. Monotube boilers have precious little residual heat in the tubing, so the boiler stops making steam, almost instantly.
Finally, before every start-up, the ECU insures all controlled valves are set to their "start" position.

In the industry we call that DDC - direct digital control.

 

[HMEL]

To avoid any confusion, my ECU should not be considered a DCS. The computer inside the ECU is an Arduino Mega 2560 microcontroller which is programed using Arduino's version of C++. There are no PIDs, PDs, or PLCs used anywhere.
Excluding software issues, if the ECU were to fail for any reason, all the electric motors instantly stop running. There would be no fuel flow and no airflow into the burner, resulting in no burner flames. Monotube boilers have precious little residual heat in the tubing, so the boiler stops making steam, almost instantly.
Finally, before every start-up, the ECU insures all controlled valves are set to their "start" position.

Bit hesitant to reply but there are several engineering issues that may be ok for smaller systems but not ok as the size goes up. Let me give you an example. Lets say your fan has a hickup like blows a fuse or has a short some where. Unless your system instantly knows this the fuel system will continue to deliver oil in those few seconds. The system trips is reset and restarts. But now with a load of unburned fuel in the furnace. That would not be pleasant. Also do not assume a monotube boiler has precious little residual heat in the tubes. Its the re-ignition probability that can cause an issue not the steam production. This is a simple fireside analysis. By the way your design is similar to a programmable logic control. But as this is an experimental design and it does what you want it to proceed until you learn otherwise.

 

[TimTaylor]

Bit hesitant to reply but there are several engineering issues that may be ok for smaller systems but not ok as the size goes up. Let me give you an example. Lets say your fan has a hickup like blows a fuse or has a short some where. Unless your system instantly knows this the fuel system will continue to deliver oil in those few seconds. The system trips is reset and restarts. But now with a load of unburned fuel in the furnace. That would not be pleasant. Also do not assume a monotube boiler has precious little residual heat in the tubes. Its the re-ignition probability that can cause an issue not the steam production. This is a simple fireside analysis. By the way your design is similar to a programmable logic control. But as this is an experimental design and it does what you want it to proceed until you learn otherwise.

Since he is using DDC, I'd suggest adding some type of an external watchdog timer circuit that has to be reset by the control board or it will cut the mains power. It's relay should be N.O. and can be initially energized when the system is started. I would never let the system auto-ignite - it should require a human to manually push the start button.

 

[Toymaker]

In the industry we call that DDC - direct digital control.

DDC,.... I'm OK with whatever acronym we can all agree upon. The only point of my post was to better inform readers of the differences between my DDC, (aka, ECU, DEC, FADEC) and a DCS, as I agree with HMEL that distributed controls come with often overlooked dangers.

 

[Toymaker]

Bit hesitant to reply but there are several engineering issues that may be ok for smaller systems but not ok as the size goes up. Let me give you an example. Lets say your fan has a hickup like blows a fuse or has a short some where. Unless your system instantly knows this the fuel system will continue to deliver oil in those few seconds. The system trips is reset and restarts. But now with a load of unburned fuel in the furnace. That would not be pleasant. Also do not assume a monotube boiler has precious little residual heat in the tubes. Its the re-ignition probability that can cause an issue not the steam production. This is a simple fireside analysis. By the way your design is similar to a programmable logic control. But as this is an experimental design and it does what you want it to proceed until you learn otherwise.

First, I appreciate you, and all others, for pointing out what you believe to be a safety or functionality issue,....as it might be something I've overlooked, or incorrectly addressed.

One of the many tasks still on my rather long to-do list for this project is to install the 0-to-15 psi pressure sensor into the leaf-blower's output air stream, (the one still sitting on workbench,...somewhere) followed by writing the code to shut off, or reduce fuel flow, whenever burner air pressure is lower than expected. The software controlling and monitoring all the various functions of this engine runs in an infinite loop, which repeats dozens of times each second, meaning a loss of airflow into the burner would be detected within a few milliseconds.

A different fuel-flame safety issue I had considered is what happens if an air or water bubble enters the fuel line resulting in the flame front being blown out, thereby allowing fuel-air mist continuing to be blown into the combustion chamber, unignited? Once a flame is detected inside the combustion chamber, the software turns the ignitor off,.... so, would the flame sensor and software respond quickly enough to prevent a dangerous amount of fuel-air mist from building up inside the boiler? Fortunately I inadvertently captured the answer on video, which shows the burner going out (due to air in the fuel line) shortly after initial start-up, and re-igniting in less than one second.

 

[Steamchick]

Some ignition systems also have a "fail-safe" (?) Ignite in the form of a hot wire or similar that has no control system, electronics or anything. Just a 2 to 3 second thermal inertia... Simly. The wire is heated in the edge of the flam at the usual ignition zone (for the optimum mixtuure) so if acdroplet of water comes through the fuel-line and the flame-out occurs, the momentary re-supply of fuel re-ignites from the hot wire. A longer pause requires a manual re-start. (After appropriate checks to avoid wet fuel or residual vapour in the system).
Not an expert view. So correct me if wrong?
K2

 

[Toymaker]

Some ignition systems also have a "fail-safe" (?) Ignite in the form of a hot wire or similar that has no control system, electronics or anything. Just a 2 to 3 second thermal inertia... Simly. The wire is heated in the edge of the flam at the usual ignition zone (for the optimum mixtuure) so if acdroplet of water comes through the fuel-line and the flame-out occurs, the momentary re-supply of fuel re-ignites from the hot wire. A longer pause requires a manual re-start. (After appropriate checks to avoid wet fuel or residual vapour in the system).
Not an expert view. So correct me if wrong?
K2

There are several ways to keep the flames burning, and your above method seems as good as any.

I use a high voltage arc between two tungsten electrodes. The High Voltage is generated by two tiny electronic boards that use minimal power to run, and all the time I spent adjusting fuel-air flow by hand to see what worked best, the ignitor was on continuously. So, I know I could just leave the ignitor on all the time, but now that I know how fast the re-ignite system works, I don't see a good reason to leave it on continuously.

 

[peterl95124]

I am currently trying to diagnose why my home heating furnace isn't working, the flame ignites but the system turns off after a few seconds, apparently the flame sensor has gone bad. Point is to be safe these ignition systems are necessarily complicated. Failure modes are hard for the human mind to anticipate, which isn't conducive to DIY hobbiest safety. IMHO, YMMV, VWPBL.

 

[Toymaker]

Getting back onto the topic of feed pumps, as I mentioned earlier, the entire steam engine, including boiler, will be installed in a vehicle, meaning it's very desirable to keep both weight and size as small as is possible.

The simplest solution feed pump would be to use a positive displacement pump driven by an electric motor, like an electric pressure washer. But electric pressure washers are both large & heavy due mostly to the 1.5 HP+ AC motor. Also, the motor being AC is another big problem for a mobile platform which cannot be plugged into an outlet. So, what other options are available?

The best solution I've come up with is to use two separate pumps. A small DC electric pump capable of delivering at least 1 LPM (Liter Per Minute) at 7 bar (100 psi) or more will be used to initially ensure boiler tubes are full of water, and to provide enough steam pressure to run the second, more powerful feed pump, which will be steam turbine powered.

Small, High Pressure fuel pumps used in many cars are smaller than a soup can, weigh only 1 kg, run on 12 VDC, and deliver 5 LPM at 7 bar. Many race car enthusiasts use two of these pumps in series to double the pressure output. Many different manufacturers make these pumps, and I've purchased 2 thru AliExpress from a company called Osias. The chart for the specific model I purchased is shown below.

If the above chart proves accurate, than two pumps in series will deliver 4.6 LPM at 15 bar (240 psi), and even using more conservative numbers from the above chart, two pumps in series should provide 5 LPM at 12 bar.

Dimensions and weight are shown below.

I should have these two pumps to test in 2 to 3 weeks, at which time I will post test results in this thread.

As for the second feed pump: I began discussing a turbine powered feed pump back in May '23 in a different thread: DIY Tesla Impulse Turbine

 

[HMEL]

I am currently trying to diagnose why my home heating furnace isn't working, the flame ignites but the system turns off after a few seconds, apparently the flame sensor has gone bad. Point is to be safe these ignition systems are necessarily complicated. Failure modes are hard for the human mind to anticipate, which isn't conducive to DIY hobbiest safety. IMHO, YMMV, VWPBL.

If you have a high efficiency furnace it can also be a pressure sensor that has water in it. This was my experience last winter. Furnace would initiate a start come on line and then drop out because of that interlock. The service technician replaced it. I dont mess with the furnace anymore due to the solid state systems on it. Beside they usually have the parts on the truck when the come to service it. However, I just replaced the heat exchanger because of a leak they found during testing. That part took about three weeks to get.

 

[HMEL]

Getting back onto the topic of feed pumps, as I mentioned earlier, the entire steam engine, including boiler, will be installed in a vehicle, meaning it's very desirable to keep both weight and size as small as is possible.

The simplest solution feed pump would be to use a positive displacement pump driven by an electric motor, like an electric pressure washer. But electric pressure washers are both large & heavy due mostly to the 1.5 HP+ AC motor. Also, the motor being AC is another big problem for a mobile platform which cannot be plugged into an outlet. So, what other options are available?

The best solution I've come up with is to use two separate pumps. A small DC electric pump capable of delivering at least 1 LPM (Liter Per Minute) at 7 bar (100 psi) or more will be used to initially ensure boiler tubes are full of water, and to provide enough steam pressure to run the second, more powerful feed pump, which will be steam turbine powered.

Small, High Pressure fuel pumps used in many cars are smaller than a soup can, weigh only 1 kg, run on 12 VDC, and deliver 5 LPM at 7 bar. Many race car enthusiasts use two of these pumps in series to double the pressure output. Many different manufacturers make these pumps, and I've purchased 2 thru AliExpress from a company called Osias. The chart for the specific model I purchased is shown below.

View attachment 151040
If the above chart proves accurate, than two pumps in series will deliver 4.6 LPM at 15 bar (240 psi), and even using more conservative numbers from the above chart, two pumps in series should provide 5 LPM at 12 bar.

Dimensions and weight are shown below.

View attachment 151041

I should have these two pumps to test in 2 to 3 weeks, at which time I will post test results in this thread.

As for the second feed pump: I began discussing a turbine powered feed pump back in May '23 in a different thread: DIY Tesla Impulse Turbine

I do not think a fuel pump has the requirements to handle water, especially hot condensate. You should ask the manufacture before you use it see what they say.

 

[TimTaylor]

I am currently trying to diagnose why my home heating furnace isn't working, the flame ignites but the system turns off after a few seconds, apparently the flame sensor has gone bad. Point is to be safe these ignition systems are necessarily complicated. Failure modes are hard for the human mind to anticipate, which isn't conducive to DIY hobbiest safety. IMHO, YMMV, VWPBL.

There are a number of things that can cause what you are experiencing. Check to make sure that both the combustion air inlet and the exhaust are clear and unobstructed. Also check the condensate drain system to make sure it is functioning correctly, as issues with condensate removal can cause pressure fluctuations in the combustion chamber that will cause the burner to shut down. The air pressure sensor could also be faulty. There are also usually 2 thermocouple sensors that are part of most ignition/safety circuits - could be a faulty sensor or bad connection.

 

[Toymaker]

I do not think a fuel pump has the requirements to handle water, especially hot condensate. You should ask the manufacture before you use it see what they say.

OK HMEL, you've peaked my curiosity,...what requirements are you referring to that would prevent the Osias fuel pump from pumping water??

Below is a drawing of the internal parts of these pumps. As you can see, they're tiny impellor-vane pumps, well suited to pump most any fluid.

My biggest question is, at what pressure will the Relief Valve open?

 

[Steamchick]

Fuel is non-conducting. Water conducts electricity. It could get lively! Is this a 12V DC motor? I can't make out the inlet (Back cover?) and outlet connections to be sure. Is this a wet motor pump? - I.E. the motor full of fluid? - Looks like it from what looks like a hose connector (Outlet?) with the valve holder and one-way valve in it? Water (pumped condensate) will not do the armature very much good. = Rust. Nor the springs for brushes, etc.
But the springs just may be stainless steel...
K2

 

[Toymaker]

Fuel is non-conducting. Water conducts electricity. It could get lively! Is this a 12V DC motor? I can't make out the inlet (Back cover?) and outlet connections to be sure. Is this a wet motor pump? - I.E. the motor full of fluid? - Looks like it from what looks like a hose connector (Outlet?) with the valve holder and one-way valve in it? Water (pumped condensate) will not do the armature very much good. = Rust. Nor the springs for brushes, etc.
But the springs just may be stainless steel...
K2

Here's a link to an Osias web page showing a much better drawing of the inside parts and assembly,...Osias 380LPH, but I still cannot see the fuel channels. Perhaps a hollow shaft??

Yes, the motor is 12V dc. So even if water does come into contact with both 12V terminals, free Hydrogen and Oxygen created via electrolysis will be limited, and will be rapidly carried away by the flow.

Rust should be minimal if an anti-rust additive is added to the water.