Forced recirculation AND feed from single pump

Big development while working on the steam mini chopper. I got the idea when thinking about feed pump bypassing...and how it is such wasted mechanical movement. I thought...instead of just pumping the feed water back to the water tank...why not make it more useful? Then it hit me...make it force recirculate when not pumping in feed water.

Ok...I still have a nervous twitch from all the problems with the LaMont lawn tractor system...but with an already high pressure, leak / problem free positive displacement pump working on the bike...that can also handle boiler temps due to its all metal construction and graphite packing...I thought "this is doable".

After sketching out a few schematics on how this could work...I figured out a way to switch the pumps input from feed tank to boiler drum water, with only a single ball valve needed to make the transition. It required only a few minor plumbing changes, a couple extra check valves and a ball valve installed at the bottom of steam drum. The feed pump circuit is essentially the same...except now the feed water is preheated before the pump. This keeps the temps from swaying too much inside the pump assembly and not getting shocked from a sudden temperature change.

This is how it works: With the ball valve closed, the pump draws its water from the feed tank, open to atmospheric pressure, just like before. The ball valve is attached to the bottom of the steam drum, and is plummed into a tee at the pumps input. When the valve is opened...the pump is now at boiler pressure. This forces the check valve at the water tank on its seat. Now the pump can only draw the carry over water from the steam drum...through the generating coils...and back into the drum. That's it. Simple, easy enough to convert and...in theory...increase steam production and boiler performance.

So....I made the necessary changes starting Saturday morning ...and had it completed and ready for testing by that afternoon. The test results met and exceeded my expectations. The system worked almost flawlessly...with a noticeable spike in steam production, pressure and engine RPMs when the valve was opened and the system was in forced circulation mode. There was a slight drop in steam when transitioning back to feed water pumping, probably due to the somewhat cooler feed water chilling the boiler. The drop was noticeable but not drastic.

I made a video of the test and posted it on my YouTube channel. I also hooked up a stainless steel chain to the centrifugal clutch and wheel sprocket, and put the bike on a stand so the back wheel can free spin. Here is a link to the video:

Jamison

[youtu.be]

Sweet, I've found that forced circulation makes everything so much easier. And you don't need as much surface area either. It's really nice, and my remotely pumped circulator worked really well until the little screws on the cylinder to adjust the cushion started leaking and allowed the cylinder to heat up and soften the printed parts. I've got a new cylinder now and I'll be dancing wooden parts to touch the cylinder. The only problem is driving the circulation before you build enough pressure to idle the engine.

Z,

I agree...forced circulation has many advantages: Greater heat transfer with less tubing, steam bubbles get “scrubbed” and don’t have time to form, less risk of overheating because tubing is flooded, etc. The main disadvantages are the need for a separate, high pressure, high temp circulating pump and the power to run it and, the dependence on the same pump for the system to work.

I believe this new system solves those disadvantages. No need for a separate pump...the simple, tried and true feed pump is relatively reliable and robust and handles the heat and pressure. It runs off the engine so no need for a high current electrical motor. It’s not needed for startup, either....the system builds up enough pressure to run the engine just by filling it with water and igniting the burner. The system is not dependent on the pump to work...a separate manually operated pump could take over if the engine driven one fails...for both feed and circ...even though it would be quite the workout. This system will also work as a once-through monotube...like before...just by keeping the recirculating valve closed.

Another advantage of this system...a sight glass or water level probe is all that’s needed to control the system, and it’s simple and easy to do manually....just close the valve to pump fresh feedwater and open the valve to stop feeding and start recirculation. This could easily be automated....I have ideas for doing just that, with no electricity involved. I also plan on replacing the brass ball valve with the 1000 deg Fahrenheit, 5,000 psi stainless globe valve I have....to better handle the volatile temps / pressures in that circuit.

Looks like a possible trial run this weekend....I know, I know....I’ve said that before...but I’m glad I took the time to design, build and test the forced circulation system, even though it set the project back a bit. I have a feeling that under full load....she is going to need that extra steam. Just need to mount a seat, propane manifold, weld front fork braces and a rear furnace guard.

Jamison

Hi Jamison,

Do you have an approximation for the circulation ratio when this is operating as a recirc. pump?

Ken

Ken,

The testing I have done so far has been non-scientific and no measurements have been taken of steam rate, water and fuel consumption or temperature. To be honest...even if I did...I wouldn’t know how to calculate the circ ratio. I don’t have an engineering background, but am a quick study....so perhaps at the upcoming meet (if time allows) those who are engineers could help me evaluate the system in such a manner, or even here on the forum, once I collect the necessary data needed for these calculations. I am curious how this system compares to a traditional LaMont and other boiler types.

Here is what I do know: The pump has a 3/8” ram with an approx 3/4” stroke. Speed of pump is 1/2 of engine speed. The tubing that is in the recirculation circuit is 50’ of 3/8” stainless and 50’ of 1/4” stainless, with a 2’ long, 2” pipe water drum between the two coils and a steam separating pipe drum of the same dimensions at the end of the circuit.

I put in fire and water....steam comes out and drives an engine. That’s what I know so far.

Jamison

Brilliant work Jamison - I really enjoy hearing your progress.
Pump speed is 1/2 engine speed - from the engine rpm determined by the chain sprocket gearing and assuming you were cruising at say 20mph you could work out a theoretical pumping rate which would presumably be similar for water feed and recirculation.

Mike

Mike,

Thanks! Hopefully, I will have the thing on the road soon.

I’ve made some changes to the system....replaced the 50’ of 1/4” tubing with about 30’ of 3/8” SS tubing. I think the 1/4” section was too restrictive and the 3/8” would flow better with the recirculation....and not flash to steam so easily in that section. With the old set up...I wanted to generate some steam in that section to create the pressure differential, to drive the column of water in the first drum through the main generating coils. I still have the 1/4” coil as it was, so swapping it back is not a big deal. I am also considering changing the 1/4” copper tube preheater to 3/8” copper tube...again for better flow and to avoid premature flashing.

I am also moving the throttle valve to after the superheater, installed right at the engine. The original “throttle” SS ball valve, before the superheater, will be kept in place as a safety shutoff. I will be fabricating a needle valve out of all 316 SS components with a graphite packing. I have a class 3000 SS 1/4” tee fitting that will make a fine valve body. The “needle” will be an unthreaded SS rod that will move linearly, actuated by the twist throttle. It doesn’t have to make a perfect seal because the engine will be idling and the valve will never be fully shut anyway. The old throttle ball valve...now a shut off valve...will cut off all steam to the engine if need be. I did this for better throttle response...before there was a good bit of volume in the superheater and plumbing after the valve and sometimes the engine would run-on after the valve was shut. This new arrangement will contribute to both safety and performance

The superheater is a vertical 1/2” stainless pipe, capped at the bottom, with a smaller tube inside it (coaxial) that forces the steam down and back up 180deg. It’s designed to trap any water slugs as well as superheat. My theory is that water was entering the superheater and it became a mini-boiler of sorts....and going straight to the engine without a way to shut it off. .

The superheater was also improved by adding an internal spiral with SS wire....the same technique Rick H came up with for his vaporizer tube. Now the saturated steam is forced in a spiral as it passes through.

I should still be on target for a test run this weekend...even with the new changes. Everything except the new throttle valve and (possible) 3/8” copper preheat tubing upgrade has been completed already. I can’t wait to take her for a spin...wind in my hair...steam clouds trailing behind me...neighbors running for their lives....

Jamison

My pump doesn't need much power at all, since it's balanced. The only torque on the motor is the difference in area for the piston rod. I've almost got it fixed up. Unfortunately I got laid off this week so my budget has become nill.