Small Scale CHP Steam Engine Project

[buenijo]

I had to stop the project abruptly after a bizarre chain of events - and not even accounting for the Covid nonsense. I was able to restart a couple months back. My budget is almost nonexistent. Therefore, it will move like cold molasses. Also, I cannot discuss many specifics on how it is configured. I apologize in advance.

GENERAL DESCRIPTION: I decided to use a propane burner to develop the test engine. I've done more than enough testing with small wood furnaces to know I can make it work when the time comes. That's among the least of my concerns. The engine is designed around a small gas engine crankcase mounted to a pump roll cage. A steam cylinder will mount to a plate that seals the top of the crankcase cylinder. The steam piston will connect to the crankcase piston with a rod sealed in a packing gland (crosshead configuration). A permanent magnet alternator will couple directly to the crankshaft PTO (1400 RPM) for battery charging via an external three phase bridge rectifier. The steam exhaust will be directed to a water-cooled condenser. The water heated by this process will be used to provide for heating applications.

PROGRESS SO FAR: A small engine crankcase is currently mounted on a pump roll cage. A gearmotor drives the flywheel end of the crankshaft via a freewheel coupling I designed. A sprocket on the gearmotor chain-drives a water feed pump. The way it works is the gearmotor drives the crankshaft. However, the freewheel coupling prevents the crankshaft from driving the gearmotor. This configuration allows the gearmotor to BOTH start the engine AND drive the feed pump at a precise speed. The motor is not stressed at all in this application.

NEXT STEPS: Mount a small oil pump to be chain-driven off the pto shaft (*). Once the oil pump is installed and tested, then I will start assembling the (propane) furnace and steam generator. Only after I can generate steam reliably (and safely) at the required rate, temperature, and pressure will I assemble the rest of the system including steam cylinder, low pressure condenser, and oil separator.

VIABLE FUELS: Wood chips take priority. But I am also interested in liquid fuels and propane. It is possible in principle to devise a furnace that can use all fuels, but this would not be easy to configure. I am confident my wood furnace design can be adapted to use propane. I may also be be able to adapt it for liquid fuels using the basic approach illustrated here: https://youtu.be/keQ83RTZ1JE?si=E1Fo1i1-zi0xbmO7&t=759. Note that I will be making use of the stock crankcase ignition system.

There is NO timeline as there are too many unknowns, and too few resources.

(*)The oil pump is to be mounted differently. See post #28.

 

[buenijo]

AC Generator

Nothing material to share. Just waiting on parts for oil pump install.

I will share an interesting configuration that is possible with my engine design. The design uses a three phase permanent magnet alternator (PMA) coupled to the crankshaft and hard-wired to a 24v (or 48v) battery via a rectifier. The engine will operate at a constant speed under this condition. Therefore, it is possible to belt drive a small AC generator head off the engine - provided the PMA is charging.

To understand what I'm talking about here, imagine the engine is operating with a PMA battery charging rate of 1 KWe. A small AC generator is belt-driven for a speed of roughly 3750 rpm. Now turn on a 10,000 btu a/c unit powered by the AC generator. This draws roughly 850 watts. The AC generator will bog down the engine taking the speed down slightly such that the AC generator is at roughly 3600 rpm for 60 hz. The PMA slows down a little to take the charging rate down to 150 watts. So the DC PMA and the AC generator share the 1 KWe load, and the a/c unit can be cycled on/off so long as the battery is charging.

I do not suggest this is a good idea, only that it will work. I say go with inverters for AC power and use DC where practical.

 

[buenijo]

Chilled Water System

Note: I discussed this concept in previous post #18. This post here provides a somewhat more detailed discussion.

An automotive compressor could be belt-driven off the engine and geared to provide a load slightly less than engine output. When the compressor engages (by energizing the solenoid clutch), then the engine bogs down with most of the engine load transferring from the PMA to the compressor. So battery charging rate goes very low when the compressor is engaged. In this approach, the compressor clutch would engage on low chilled water temperature. The fans on the fan coil units are thermostatically controlled. So, fan speed increases as temperature in the home rises, and vice versa - or the fan speed can be adjusted manually just like a typical automotive a/c system.

The hardware required for this system is not expensive. Also, it's not terribly complicated mechanically. For example, the compressor would operate at a constant output with more or less constant conditions, and so a needle valve or fixed orifice could be used in lieu of a TXV, and no accumulator is required to store excess refrigerant. It might make sense under the right set of conditions. A side benefit here is, while more fuel is required when using the engine for a/c (as compared to using a large solar array to power a compressor motor), there would be A LOT of heat available at the steam condenser that could be used for wood drying. Also, in principle, the same fan coil units used for hydronic space heating could be configured for chilled water to provide a/c.

ADDENDUM: Automotive a/c systems have high cooling capacity because cars are basically greenhouses on wheels. They also have to be designed for regions like Phoenix, AZ during summer. A typical automotive a/c system in a sedan is capable of a cooling rate on the order of 2 tons. Unfortunately, the efficiency is poor because the heat exchangers are relatively small. For this reason, the temperature differentials must be much higher to achieve the required heat transfer rates. This means the differential pressure across the compressor is much higher than a residential a/c setting. In fact, the compressor can put out a lot more refrigerant than the heat exchangers can handle. Therefore, automotive a/c systems have a high pressure switch that will de-energize the clutch if the compressor discharge pressure exceeds a certain value. However, a typical automotive compressor could support a 5 ton a/c system in a residential a/c setting.

An automotive compressor used in a residential a/c system would show a pressure differential roughly half that seen in the automotive setting. Therefore, the power required for the same refrigerating capacity would fall by roughly half. Furthermore, the higher evaporator temperature and pressure also implies a higher refrigerating capacity at the same rpm (the density of the refrigerant in the evaporator would nearly double). Therefore, at the same compressor speed, the cooling rate would nearly double. Lower compressor speed and lower differential pressure also implies less wear.

SOME CONSIDERATIONS: The volumetric efficiency of automotive piston compressors is about 55-65% (with the highest values at lower speeds). The density of R134a saturated vapor at 40F (suitable for a chilled water system) is about 1.05 lb/cf. The latent heat to vaporize a pound of this refrigerant is about 85 btu. I used this information to estimate the required rpm as a function of displacement and refrigeration capacity. A popular Sanden model is 138 cc (about 8.4 ci). My figures show approximately 710 rpm for a one ton refrigeration capacity (it would be about twice this speed in the automotive setting). This changes in direct proportion to the rpm. The required shaft power for this refrigeration capacity under automotive conditions is listed by Sanden as about 2.6 hp. A conservative estimate for the required power with a chilled water system is about half at 1.3 hp. A water-cooled condenser could be driven down to much lower temperatures (and pressures) to take the required shaft power to less than 1 hp.

NOTE: Many of my posts are educational. While the previously described configuration will certainly work, I consider it superior to use conventional a/c systems powered primarily by photovoltaics. Solar power + air conditioning is a powerful paradigm because the sun is often shining when it's hot. It just makes sense.

 

[buenijo]

I did get some parts for the oil pump install. But after tinkering a while, I decided to modify the design. This is an opportunity to explain part of my process here. The design philosophy centers on repurposing readily available low cost parts. It takes a lot of time to source the right parts - and ideas don't always work the first time. This is pretty much how all projects unfold. However, the difference here is I often come across near perfect components that are simply too expensive. I decline not (primarily) because I lack the means, but because the goal is to finally devise a system that can be produced cost-effectively at modest scale. Therefore, parts that are not easy to come by or otherwise costly do not pass muster. Other times, the parts are low cost, but the required modifications are too onerous (i.e. labor costs too high). It's frustrating, but it's the only way I see this working in the long run.

While tinkering with the oil pump, I did a series of tests on the water feed pump. I operated the gearmotor at a fixed speed that fell within the engine design specs. I then made a series of five test runs each exactly 60 seconds and carefully measured the volume of water using a graduated cylinder. The volume came to precisely 165 ml on each test. The ability to maintain a constant water flow rate into the steam generator is important.

I also did a cursory test with a partial vacuum on the pump suction. This reduced and otherwise varied the flow rate (as expected). The implication is the condenser pressure must be maintained above a certain value - and this is not a problem because the system is designed as a CHP unit meaning a higher condenser temperature (and pressure) is inherent in the design. That noted, it's also possible to provide a very small booster pump, or place a vacuum breaker to keep the pressure above a minimum value. Driving the condenser pressure as low as possible will improve efficiency measurably. But I really don't care because HEAT is the primary value here. The single most important design goal is RELIABILITY with the ability to easily make field repairs a close second.

 

[buenijo]

Minor progress (I'll take what I can get). I found a low cost oil shaft seal for the oil pump shaft extension I installed. It seems to work well. Also, the damned barbed fittings I installed leak at the threads (chinese seem to have trouble with npt threads - among other things). I couldn't find crush washers of the right size, so I sealed the threads with jb weld. I'm quite certain this will work. The oil pump seems to have a higher flow rate than I expected. I may have to reduce the speed with a larger sprocket. Not a big deal. Looks like it's gonna work. I'll provide a pic after install. Gonna be a few weeks.

ADDENDUM: The oil pump is to be mounted differently. See post #28.

 

[buenijo]

Free fuel. Delivered.



https://youtu.be/CFpdjnZi_5Q?si=iSzXHAWqcIowoMxi

https://youtu.be/E2IGRw4MBcE?si=ko5tzHOz1Jq3JbN-

"ChipDrop matches people who want free wood chip mulch with arborists and tree companies who are trying to get rid of it.

By signing up and placing a request you'll be added to a list of people in your neighborhood who are trying to get free wood chips.

The next time a local tree trimming company is in your area, they might deliver a load to you."

 

[buenijo]

Steam Generator Test Unit Description

The steam generator will mount to the roll cage. When the test engine is assembled, the steam line will connect to the steam cylinder head. However, for steam generator testing, the steam line will connect directly to a SS tubing coil submerged in a water bath. I will be using my 8 gallon still to contain the water.

The still has a 1500 watt heating element installed in the side. I will be removing the heating element and installing in its place a fitting which connects to the condenser coil on the inside while mounting a needle valve and pressure gage on the outside. I'll be setting the pressure by positioning the needle valve. The way this would work is high pressure superheated steam would enter the condenser coil and condense the steam at full steam generator pressure (600-800 psig). The steam will fully condense at saturation temperature which on the order of 500F in this pressure range. Therefore, the water in the still will BOIL because it receives ALL the latent heat from the steam(*).

NOTE: My still will also be used for engine testing. When the time comes, I will install a heating element for loading the engine-driven alternator directly.

(*)I completed this test set up in early January of 2025 and tested several times. All the superheated steam is easily condensed at full steam generator pressure. The water leaves the system through the needle valve. Throttling the needle valve sets the pressure. With the propane burner constant, I can set the steam line temperature by adjusting the feed pump speed. It works perfectly (see post #37).

 

[buenijo]

New Way to Install the Oil Pump

The oil pump works, but I arrived at a much better way to install it. I had removed the engine camshaft because I simply do not need it. However, I considered that I can use it to drive the oil pump directly. This will eliminate two sprockets, a chain, and the angle bracket - along with a need for a chain guard and chain lubrication. Also, the camshaft operates at 1/2 crankshaft speed. Perfect. The oil pump will operate at only 700 rpm.

I drilled and tapped the camshaft and threaded in small SS threaded rod that extends out of the crankcase cover. Eventually, an adapter plate will mount on the crankcase cover for mounting the alternator. I can then mount the oil pump on this adapter plate and couple the shaft to the M6 threaded rod extension on the camshaft.

 

[buenijo]

Next Steps

I put the oil pump in storage. I won't be dealing with that again until the alternator is ready for installation as the oil pump will mount on the alternator adapter plate - and this will not take place any time soon.

Next step is to start making fire. I need to acquire the propane burner and start assembling a suitable combustion chamber.

The goal is to establish a high temperature clean burn at a controlled rate of up to 15 KW. After this point, I will form the steam generator coil and start building a test stand. I'll start assembling the steam expander only after I can safely and reliably generate steam at the required rate, temperature, and pressure.

 

[buenijo]

Update

(06/02/24) I'm setting things up to restart the project after sustaining a serious injury about 6 months back that required surgery. So far it seems I am recovering well. Although, I have some limitations. NOTE: This injury had nothing whatever to do with the project btw.

The next step is to assemble the steam generator. Almost everyone who thinks about steam engines focuses on the "steam expander". Yes, this is important for good results. However, the expander is relatively simple compared to the rest of the system. Fortunately, I have my steam expander design worked out with all parts sourced. Hopefully, it will come together quickly once the steam generator is finished. I have a nice table on casters to mount the test unit. It's just a matter of figuring out how to bring it all together.

(06/13/24) This was a spur of the moment brainstorm that I cobbled together in a few hours to roll the steam generator tubing coil (pic deleted). The 1/4" stainless tubing here has a very thick wall at 0.049", so it's damned hard to work with. We'll see how my simple rig works.

(06/15/24) The simple rig worked surprisingly well! If I ever have to roll these in quantity, then it would be easy to install a gearmotor. It was quite the workout ratcheting that thing for nearly an hour.

 

[buenijo]

I ordered a second coil of tubing with a somewhat thinner wall at 0.035" for a ridiculously low price (I couldn't pass it up). I'm gonna see if I can make a better coil on the second go. BTW, I got the thick-wall tubing coil for FREE (which is why I used it). I got lucky with the second coil. Again, my budget is peanuts.

(06/18/24) Just ordered the steam generator casing and end caps. Still waiting on the new tubing coil.

(06/19/24) New tubing coil arrived.

(06/21/24) I rolled the new tubing coil. Honestly, it did not turn out any better (or worse) than the previous coil. But I'll use the new coil b/c it's lighter. I was able to roll it much faster the second time. Again, this tube rolling rig would be very effective with a gearmotor.

(06/22/24) I received the steam generator casing - an 8" diameter welded SS duct 24" long (good quality and GREAT price!). Gonna cut off about 5". The layout of components on the roll cage is good. For example, the feed pump discharge fitting is right next to the base of the steam generator and about 1" below it. So, the water end of the steam generator coil needs to extend only a few inches outside the base of the steam generator for connection to the feed pump.

(06/25/24) I have most of the materials required to install the steam generator. I've got a reasonably good plan for assembly. But I have to figure it out as I go, and that takes time. I often have no days available during the week to work on the project - and sometimes only a few hours.

 

[buenijo]

I decided to re-roll one of the previous coils using a smaller diameter form. This should compensate for the springback I mentioned previously. We'll see how it goes. The original design requires the coil diameter to be 1/2" smaller. This should not delay the project in a material way. In fact, it should make the final installation a lot simpler.

(07/11/24) I re-rolled the tubing coil, and it turned out well. Definitely an improvement. This will absolutely make the steam generator assembly easier. There also is now room for nearly a full inch of ceramic fiber insulation on the casing.

(07/13/24) Ordered 6" SS stove pipe to contain the casing insulation and shroud the steam generator tubing coil. Good price.

(07/28/24) The steam generator is assembled! I confess I still need to make the casing cap and removable plug, but the hard part is done. I monkeyed around with it for a while before I figured it out. It was surprisingly simple to assemble. I hope it works because with practice I could throw these together quickly. Of course, more parts and work is required to assemble the full test rig. I have to shape the condenser coil, get some fittings, etc.

(08/01/24) Steam generator casing cap is installed including insulation. I also installed the removable plug. It turned out good. I now need to install a fairly heavy steel plate on the underside of the plug. It will extend 2" below the plug using a bolt with this 2" gap insulated. I placed a very heavy steel plate on top of the combustion chamber and tubing coil. This very heavy steel plate has a 3" center hole. The purpose of the steel plate on the plug is to cover this center hole. When the plug is in place, the combustion gases move through a 1/2" gap between the top of the combustion chamber and the base of the heavy steel plate, then down across the steam generator tubing coil. However, when the plug is removed, the gases vent out the top of the steam generator casing.

(08/10/24) The removable plug is installed in the steam generator cap. So, the steam generator is completed. I dragged out the still, cleaned it up, and started purchasing parts for the test rig. I have a good needle valve on hand, and I'm going to use the first tubing coil I made for the steam generator as a condenser coil. I have to cut it in half to fit it in the base of the still (so, roughly 50 feet of 1/4" SS tubing). This will be more than enough to condense the steam in this setting (water-cooled and with a temperature differential of nearly 300F).

 

[buenijo]

I now (almost) have all the parts for the steam generator test rig. Just have to find the time to assemble everything. I received two high temperature digital thermostats with K probe this morning. I installed a probe in a 1/4" copper tee welding joint and cut the tee such that it clamps to the 1/4" steam line. The steam line and the tee will eventually be insulated with three layers of fiberglass sleeves (which I have on hand). I tested the thermostat by heating the steam line with a torch about an inch from the tee. I took it up to 300C. It was reasonably responsive. It also seems accurate based on brief testing. If anyone is interested in a low cost and seemingly decent high temperature thermostat, then see the attached video. Here is a link for specs and purchase: https://www.mpja.com/High-Temperature-Thermostat-30-to-999Deg/productinfo/34687 MP/. BTW, replacement probes can be found on Amazon for about $3 each.

I am using the thermostat only for temperature indication for steam generator testing. But I may use the relay for engine testing. It has both normally open and normally closed configurations. Great deal too - I got two units for less than $40. The product is an exposed circuit board, so I installed it in a clear plastic case and drilled holes for the wires to penetrate.



(08/29/24) High pressure steam condenser coil is installed in the pot still and connected to needle valve. Gotta connect steam line and condensate line, get a few more odds and ends, and set it up.

(08/31/24) Received a small 24v mag drive water pump and tested it. Impressive little pump! It weighs only 1/2 pound and is less than 3" long, but pumped 2.5 gpm. Max head pressure is listed at 24 feet, but I didn't test that parameter. Gonna use this to pump cooling water through the plate heat exchanger(*). Oh, and the specs claim it can tolerate water temperature up to 200F. I need up to about 160F. Also got some nice electrical connectors for all the components, a better PWM controller for the feed pump, and a 12v power supply for the thermostat. Nothing big, just a few more baby steps.

(*)I wound up using the pump as a booster for the feed pump. Originally, I was going to send the condensate leaving the condenser through the heat exchanger to cool it down completely. But it is easier to just collect the hot water and let it cool (using it again later to prevent wasting water). This condenser is very simple and highly effective.

 

[buenijo]

HEAT ONLY Engine Configuration

My engine design can be configured to provide for heating applications without running the engine. This can be done by running the feed pump in reverse, and opening a steam bypass valve at the steam cylinder head. However, a superior approach is to place a SECOND small bypass valve at the feed pump discharge. The way this works is, as the condensate in the steam generator vaporizes, the head pressure at the base of the steam generator falls, and this allows the condensate from the condenser (which is elevated) to replenish the water in the steam generator. In this way, the system can generate low pressure steam for heating applications without running the feed pump.

A side benefit is the ability to service (even remove) the engine components without disrupting heat production. This would be done by disconnecting the steam generator at both ends, and installing jumper hoses to and from the condenser.

The average U.S. household uses 18% of annual energy consumption for water heating, and 42% for space heating. Clothes drying makes up 4%. That's already at 64% of total energy as heat. Of course, a home in a cold region will see a higher figure (76% for Michigan). In short, the ability to operate a small CHP engine system in the off grid setting for heat without having to run the engine seems like a valuable feature. Why put run time on the engine when solar can provide electricity? Furthermore, if heat demands are high and electricity demands low, then a charged battery can carry the modest electrical loads. No need to run the engine during this period. The engine can be quickly reconfigured for battery charging when required.

 

[buenijo]

DC Electric Wood Chipper & Chunker

Just ideas:

CHIPPER: Remove the engine from a small wood chipper. Install an adapter plate and mount a brushed DC permanent magnet motor. The motor I have in mind is only around 2000 rpm at 24v (but capable of twice the rpm - and power - at 48v). The peak torque of this motor is more than twice that of the stock wood chipper engine and the torque increases as the chipper bogs down. In my opinion, the lower speed at 24v is an advantage for less shear force on the shaft. Although, the motor shaft is thicker than the stock engine shaft.

CHUNKER: A relatively small DC gear reduction motor used to drive a small gearbox for further gear reduction could generate significant torque for powering a wood chunker suitable for branches up to 2" - and possibly more. I like this alternative because it would be A LOT more compact - even small enough to mount to the frame of the engine itself.

 

[buenijo]

Just making a recap post - mainly for my benefit to organize my thoughts.

RECAP (09/05/24): Very little time available. I'm nearly setup to start testing the steam generator. Again, the test rig is designed to generate high pressure steam and immediately condense the steam under full steam generator pressure (roughly 600-800 psig and 500F). This will be done by connecting the steam generator to a condenser coil placed in my pot still and submerged in a water bath. The boiling water bath will remove the vast majority of the heat.

(10/15/24) I'm almost set up for testing. But a few devils are popping up. The propane burner is extinguished when the steam generator cap is placed. This was expected. So, I had already built an attachment to the burner out of ABS pipe that included a small blower. This proved to solve the problem. However, the blower is a bit underpowered. So I installed a more powerful unit. The feed pump for this test unit is chain driven off the gearmotor. I was never happy about the size of the sprocket. So I did a quick search and found the perfect sprocket for a good price. Now I must replace the chain because, well, the new sprocket is designed for a different size chain (a good thing as the other one was overkill). The devils are in the details.

(10/20/24) I replaced the chain for the pump drive. Oddly enough, the pump decided to stop pumping after I made the switch. I figured out the problem, and I am working to resolve it. The needle valve I am using to put back pressure on the pump is a bit large. A smaller valve would work better. Also, I need to install a needle valve in the propane line to improve burner control.

(10/23/24) FAMILY DRAMA (again). The project will be further delayed. My mother-in-law's caretaker just dropped dead of a heart attack. My wife and I have taken her in. She has severe dementia and requires more or less constant care.

(12/01/24) Very little time for the project (we're still caring for my mother-in-law). But I managed to solve the ongoing feed pump problem. I tested the pump several times at pressure. I can now vary the flow rate and discharge pressure as required. I will do more testing. But I'm confident the problem with the feed pump is rectified.

(12/07/24) I reconfigured the propane burner. It's a more elegant design and with a stronger blower that includes speed control. Unfortunately, the wife sold our grill (without telling me!), and the deal included my propane tank. So, I ordered a new propane tank. I also ordered a needle valve to better control propane flow to the burner.

(12/15/24) Received new propane tank and filled it. Also received and installed the needle valve. Still need to test.

(12/20/24) Made a brief test of the new burner and blower. Seems to work well. I can now precisely vary the propane flow rate, the blower fan speed, the feed pump flow rate, and the system pressure. Next is a lengthy test to set the burn rate. Once the right needle valve setting and blower rpm is determined, then these settings will be held constant for testing. The burn rate will be held constant during testing. The pressure will be set by a needle valve on the condenser discharge.

 

[buenijo]

I tested the steam generator on 01/03/25. I was able to generate superheated steam at 400-600 psig and 600-700F. Unfortunately, the steam line fitting started leaking after a while. I have reason to believe that I installed that fitting improperly. I have additional ferrules. So I'll fix that - and do it correctly this time. Also, I need to change the plug that seals the top of the furnace combustion chamber as it's leaking a bit. NO leaks on the water end. I expect to be working with this for a several months at least before I feel confident enough to start assembling the steam expander.

 

[ClaytonB]

I tested the steam generator on 01/03/25. I was able to generate superheated steam at 400-600 psig and 600-700F. Unfortunately, the steam line fitting started leaking after a while. I have reason to believe that I installed that fitting improperly. I have additional ferrules. So I'll fix that - and do it correctly this time. Also, I need to change the plug that seals the top of the furnace combustion chamber as it's leaking a bit. NO leaks on the water end. I expect to be working with this for a several months at least before I feel confident enough to start assembling the steam expander.

Looking forward to future updates. Would love to see photos/video, as well as plans/diagrams (explanation how others can do it). I feel like we might have a lot more steam in our future...

 

[buenijo]

Looking forward to future updates. Would love to see photos/video, as well as plans/diagrams (explanation how others can do it). I feel like we might have a lot more steam in our future...

Hi Clayton. I doubt pics will be useful as this is a test rig designed to facilitate development. A final configuration would be very different. The main goals for the test rig are to develop the steam generator and steam expander. If these prove out, then I would work on a final wood furnace design. The crude wood gasification furnace depicted in the OP was just a proof of concept. I would also assemble a small plunger pump to be driven by the camshaft. The goal at that point would be to assemble a complete wood-fired steam power plant prototype unit.

 

[buenijo]

I tested the steam generator again on 01/10/25 after replacing the steam fitting and modifying the plug on top of the combustion chamber. It took nearly 10 minutes to start generating superheated steam. There was a lot of hysteresis (temperature was all over the place - but no higher than 800F). Once I got it dialed in, it held pressure from 600-650 psig with steam line temperature of 650-700F for approximately 30 minutes. Unfortunately, a rubber attachment on the propane burner (used to direct air from a blower fan into the burner head to overcome back pressure from the steam generator) melted a bit and I saw some flames escaping. So, that's a good reason to shut down. I may try sealing with silicone tape or rtv.

Based on testing, it seems like the combustion chamber is too small to achieve my originally desired power (see next post, I concluded the problem is insufficient air to the burner).