rolphill
Well-Known Member
Alternate title: Propane Tank Boiler (clickbait)
So I'm sort of happy with the boiler I built last year. It makes plenty of steam for its size, but managing the fire is more time intensive than I expected, and it only provides enough steam when the fire is just right. It's probably going to be rehomed into my boat this summer, where I think it'll do well. However, it was originally intended to power a trackless locomotive, which it just ended up being unsuited for. The fire door was so far down that it made the ergonomics unreasonable, which compounded the finnicky fire management. It was also rather underpowered for tasks involving rubber tires, gravel roads, grass, and hills. I don't want mini traction engine speeds, I want to cruise! I've dreamt of driving my own steam car since I was a teenager, and this boiler is not going to take me there. It's time to accept it and move on to something bigger and better, so let's get started!
First, the design goals. I need more boiler horsepower, like at least 3-5. I don't have $5000 to spend on having a large pressure vessel made, although getting a small tig welder isn't out of the question nowadays. It must be easy and cheap to build (for my skills and tooling). I also want a big firebox with a big door that can fit actual wood in it, and it will probably have to be horizontal so I can actually reach the firebox. Also, one guideline that heavily inspired this specific design was no pancake coils, because without special equipment you always end up with a big hole in the middle that needs to be blocked. I want a bit more pressure this time, so I'm shooting for 125psi, which is about the limit before everything gets really expensive. All tubing should be either 1/4 or 3/8in, because I can get steel brake line cheaply in those sizes, and I want to avoid copper tubing.
Alright, let's get down to the design concept. Here it is so far.
It's a horizontal locomotive boiler shaped object! To start, the firebox is a 20lb propane tank. There will be a firedoor in the back, and the entire inside is lined with coils (blue). The superheater coils (orange) are in the front of the firebox. Then, in place of a boiler barrel is a square section with 4 sets of transverse coils (yellow and red) would so they can fit inside each other. The top sheet of this square barrel section will be bolted on for easy access to the coils. This will be covered by a round "fake barrel" filled with insulation. The four flue coils are followed by an economizer coil (blue) and a smokebox and chimney. I will probably add a blast pipe as well. All in all this is about 10-12sqft of heating surface.
The inspiration that kicked off this design was this monotube boiler. I had no idea why I had never thought of this coil arrangement before, but its discovery opened up a world of possibilities (as long as you have rectangular enclosures). Two coils wound reverse to mesh with each other and placed sideways are super easy to make and provide the right amount of restriction without leaving open passages for flue gas to skip by. All five barrel coils will be installed from the top, with their ends going through holes in the bottom of the barrel box. This will allow them to easily be maintained or replaced if necessary.
This will not be a monotube boiler, it's going to be a lamont forced circulation boiler. I've experimented with the concept a couple times before. I'm going to use another one of the vessels I built my last boiler out of as the steam drum. This has a disadvantage that I'll talk about later, but it's functionally free compared to the cost of building a steam drum. Also this drum will be unfired, for what it's worth. I'll also use the same design concept for the circulation pump, but with at least two of them. The circulation pump is made from a gear motor that drives an industrial air cylinder (or two). Both ends of the air cylinder are hooked up to tees with check valves on either side to form positive displacement pumps. The connection is purposely made with a long piece of uninsulated 1/4in tubing to keep the heat away from the cylinder and motor. Since both sides of the (double acting) cylinder are under boiler pressure, the effort to pump is minimal in both directions. The motor turns at around 60rpm, which minimizes cavitation. The check valves are off the shelf all bronze gravity flap valves, for minimum cracking pressure. Menards actually sells 125psi steam rated check valves for <$10! I want to have at least two circulation circuits, the firebox, and the barrel coils.
I'm planning on using two (redundant) rv water pumps for feed water, and automatic water level control with manual override. Each one should be able to feed the boiler individually at max fire. Did you know that Aliexpress sells boiler water level electrodes with npt threads for a pittance? It certainly feels more professional than model T spark plugs. They will share the economizer. I plan on putting all of these mechanisms underneath the barrel.
Finally, some stats:
Heating Surface:
1sqft economizer
5.3sqft firebox generating coil* note the outside half of this coil is facing the wall and part of it is underneath the grate, but there will be gaps between coils.
3.6sqft barrel generating coils
1.1sqft superheater coil* same as the firebox coil
Flow rates:
Now, lamont boilers are supposedly a bit special. It's said that they can have boiler surface rates as low as 1.5sqft/bhp, but there isn't much experimental data to go on. I chose an estimate of 2sqft/bhp (16,738btu/sqft) based on two sources. This website states that the doble steam car achieved up to 26,000BTU/sqft, while this website tested their lamont boiler and achieved 9,620BTU/sqft. I doubt I'll ever get the number I chose for various reasons, but I figured it's a good estimate to ensure I conservatively overestimate pump sizes and the like. I can always place a speed control on the circulation pumps.
There is one number that is concerning though, the time to low water. This is an estimate of how long it takes to boil off the effective water level volume at full heat input. The steam drum I'm using is only 3.25in x 10.9in, for a total volume of less than 1.5L. That doesn't include space for baffles and water separation, and the fact that the gauge glass starts an inch or two above the bottom. I figured half a liter for an "effective" volume of water. However, this means my gauge glass will go from full to empty in less than 30 seconds at full fire. I know it'll be twice that at best in real life due to the solid fuel depending on draft and the conservative heat estimations, but it still seems quite fast. In any case, the steam drum is completely independent from the rest of the boiler, and can be changed out for something bigger if necessary. I also wonder about the 50% pressure -> blowoff time, but calculating that would require taking into account both the water in the tubing and the metal of the tubing and everything, and probably wouldn't be all that accurate anyways.
Other than that, the whole thing seems relatively straightforward. The tubing should only be about $100, and I already have most if not all of the sheet metal required. I've already got the valve off the chosen propane tank, so it's just waiting for a warm day so I can fill it with water (very important) and cut it open (not my first rodeo with this). After which I'll start taking pictures.
A small question:
Is there a cost effective alternative to standard brass flare fittings? I'm having a hard time finding stainless steel flare fittings. It's really only a concern when joining tube sections inside the firebox, but it would be nice to get away from brass without doubling the budget. Or maybe I should just stick with brass due to SS corrosion cracking?
So I'm sort of happy with the boiler I built last year. It makes plenty of steam for its size, but managing the fire is more time intensive than I expected, and it only provides enough steam when the fire is just right. It's probably going to be rehomed into my boat this summer, where I think it'll do well. However, it was originally intended to power a trackless locomotive, which it just ended up being unsuited for. The fire door was so far down that it made the ergonomics unreasonable, which compounded the finnicky fire management. It was also rather underpowered for tasks involving rubber tires, gravel roads, grass, and hills. I don't want mini traction engine speeds, I want to cruise! I've dreamt of driving my own steam car since I was a teenager, and this boiler is not going to take me there. It's time to accept it and move on to something bigger and better, so let's get started!
First, the design goals. I need more boiler horsepower, like at least 3-5. I don't have $5000 to spend on having a large pressure vessel made, although getting a small tig welder isn't out of the question nowadays. It must be easy and cheap to build (for my skills and tooling). I also want a big firebox with a big door that can fit actual wood in it, and it will probably have to be horizontal so I can actually reach the firebox. Also, one guideline that heavily inspired this specific design was no pancake coils, because without special equipment you always end up with a big hole in the middle that needs to be blocked. I want a bit more pressure this time, so I'm shooting for 125psi, which is about the limit before everything gets really expensive. All tubing should be either 1/4 or 3/8in, because I can get steel brake line cheaply in those sizes, and I want to avoid copper tubing.
Alright, let's get down to the design concept. Here it is so far.
It's a horizontal locomotive boiler shaped object! To start, the firebox is a 20lb propane tank. There will be a firedoor in the back, and the entire inside is lined with coils (blue). The superheater coils (orange) are in the front of the firebox. Then, in place of a boiler barrel is a square section with 4 sets of transverse coils (yellow and red) would so they can fit inside each other. The top sheet of this square barrel section will be bolted on for easy access to the coils. This will be covered by a round "fake barrel" filled with insulation. The four flue coils are followed by an economizer coil (blue) and a smokebox and chimney. I will probably add a blast pipe as well. All in all this is about 10-12sqft of heating surface.
The inspiration that kicked off this design was this monotube boiler. I had no idea why I had never thought of this coil arrangement before, but its discovery opened up a world of possibilities (as long as you have rectangular enclosures). Two coils wound reverse to mesh with each other and placed sideways are super easy to make and provide the right amount of restriction without leaving open passages for flue gas to skip by. All five barrel coils will be installed from the top, with their ends going through holes in the bottom of the barrel box. This will allow them to easily be maintained or replaced if necessary.
This will not be a monotube boiler, it's going to be a lamont forced circulation boiler. I've experimented with the concept a couple times before. I'm going to use another one of the vessels I built my last boiler out of as the steam drum. This has a disadvantage that I'll talk about later, but it's functionally free compared to the cost of building a steam drum. Also this drum will be unfired, for what it's worth. I'll also use the same design concept for the circulation pump, but with at least two of them. The circulation pump is made from a gear motor that drives an industrial air cylinder (or two). Both ends of the air cylinder are hooked up to tees with check valves on either side to form positive displacement pumps. The connection is purposely made with a long piece of uninsulated 1/4in tubing to keep the heat away from the cylinder and motor. Since both sides of the (double acting) cylinder are under boiler pressure, the effort to pump is minimal in both directions. The motor turns at around 60rpm, which minimizes cavitation. The check valves are off the shelf all bronze gravity flap valves, for minimum cracking pressure. Menards actually sells 125psi steam rated check valves for <$10! I want to have at least two circulation circuits, the firebox, and the barrel coils.
I'm planning on using two (redundant) rv water pumps for feed water, and automatic water level control with manual override. Each one should be able to feed the boiler individually at max fire. Did you know that Aliexpress sells boiler water level electrodes with npt threads for a pittance? It certainly feels more professional than model T spark plugs. They will share the economizer. I plan on putting all of these mechanisms underneath the barrel.
Finally, some stats:
Heating Surface:
1sqft economizer
5.3sqft firebox generating coil* note the outside half of this coil is facing the wall and part of it is underneath the grate, but there will be gaps between coils.
3.6sqft barrel generating coils
1.1sqft superheater coil* same as the firebox coil
Flow rates:
Now, lamont boilers are supposedly a bit special. It's said that they can have boiler surface rates as low as 1.5sqft/bhp, but there isn't much experimental data to go on. I chose an estimate of 2sqft/bhp (16,738btu/sqft) based on two sources. This website states that the doble steam car achieved up to 26,000BTU/sqft, while this website tested their lamont boiler and achieved 9,620BTU/sqft. I doubt I'll ever get the number I chose for various reasons, but I figured it's a good estimate to ensure I conservatively overestimate pump sizes and the like. I can always place a speed control on the circulation pumps.
There is one number that is concerning though, the time to low water. This is an estimate of how long it takes to boil off the effective water level volume at full heat input. The steam drum I'm using is only 3.25in x 10.9in, for a total volume of less than 1.5L. That doesn't include space for baffles and water separation, and the fact that the gauge glass starts an inch or two above the bottom. I figured half a liter for an "effective" volume of water. However, this means my gauge glass will go from full to empty in less than 30 seconds at full fire. I know it'll be twice that at best in real life due to the solid fuel depending on draft and the conservative heat estimations, but it still seems quite fast. In any case, the steam drum is completely independent from the rest of the boiler, and can be changed out for something bigger if necessary. I also wonder about the 50% pressure -> blowoff time, but calculating that would require taking into account both the water in the tubing and the metal of the tubing and everything, and probably wouldn't be all that accurate anyways.
Other than that, the whole thing seems relatively straightforward. The tubing should only be about $100, and I already have most if not all of the sheet metal required. I've already got the valve off the chosen propane tank, so it's just waiting for a warm day so I can fill it with water (very important) and cut it open (not my first rodeo with this). After which I'll start taking pictures.
A small question:
Is there a cost effective alternative to standard brass flare fittings? I'm having a hard time finding stainless steel flare fittings. It's really only a concern when joining tube sections inside the firebox, but it would be nice to get away from brass without doubling the budget. Or maybe I should just stick with brass due to SS corrosion cracking?
