Shay Operation - More Improvements
Water Hose Connectors & Better Fire Pan
Nelson Riedel
Initial: 09/23/05 Last Revised: 10/23/2006


The shay is near the end of the second summer operation and for the most part it has run very well.  Most of my time and energy are now spent on the design and construction of the Heisler.   Much of the Heisler design is based on the Shay and as such I frequently ask myself is there is a better way.   This has lead me to go back and make a couple changes in the Shay, one is the use of quick connects on the water hoses ( a minor change with a major convenience improvement) and a redesign of the fire pan  which significantly improved the performance.  

Water Hose Quick Connects:  The water hose connections under the tender used 1/4" OD tube compression fittings with the tube shoved into the end of the hose.   These worked great most the time.  However, once in a while I over tightened the nut.   When I then tried to loosen the nut the fitting or the nipple through the tank floor started to unscrew.   I then had to get out a second wrench and try to hold the compression fitting while unscrewing the compression nut.   One day when out at the track I noticed one of the guys had small quick connects in the water lines.  I asked him about them and he said they were hydraulic quick connects.   Later I asked Dick McCoy what he used and he said he used the small hydraulic quick connects and also used a small coolant quick connect.  The coolant quick connects however don't seal the line when disconnecting.   He said McMaster-Carr carries several types of quick connects.

When I got home I went to and searched on quick connect fittings and found about a dozen different types.   I ended up selecting the coolant type Dick suggested.   The photo  shows the two piece unit.  The male part (stud) has a 1/4" NPT thread.  I used a 1/4" to 1/8"  NPT bushing to match with the 1/8" nipple through the tank floor.  The 1/4" ID water hose mated exactly with the barb fitting. 

This photo shows an installed and connected quick connect.  That is the park brake lever in front of the quick connect.

When the fitting is disconnected, the tank drains though the stud which eliminates the need for a separate drain valve.  Next time I'll use a 1/4" NPT nipple through the tank bottom to increase the flow when draining the tank.  

New Fire Pan:    The burner ran pretty well.  There were however several things I wished were different:

  • It was impossible to maintain a low heat level.  At the lowest setting the safety valves would be  relieving when running slow on a flat track.

  • I seemed to need too much blower on high settings to keep from smoking (a smoking oil fire caused the tubes to collect soot quickly).

  • The spark plug didn't seem to relight the fire as easily as other oil burners I'd observed.    

I was working on the Heisler design and was investigating various superheater alternatives.  I was leaning heavily toward a heat exchanger in the top of the firebox.  Ken Schroeder advised that his experience was that smoke box superheaters didn't work very well.  (He said he'd had good results with superheater tubes in the boiler tubes and fair results with the superheater tubes in the firebox.) 

Being an engineer, I decided it was wise to try to obtain some data.   Over Labor Day Dan Staron was running his propane fired Shay and I had my oil burner on the track.   (We did a shay pull to see which was the more powerful.  That wasn't any fun; they seemed to be exactly matched)   I used a little IR thermometer to measure a few temperatures after the shays had been operating a while at ~ 100 psi.  The non insulated  boiler in the firebox area measured about 340 degrees (F) as expected.  The smoke boxes were hotter and the maximum temperature was around the base of the smokestack.  My oil burner measured about 460 and Dan's propane burner measured about 490.  

I figured the hotter the smoke stack, the hotter the flue gases and the better a smoke box superheater would work. So, later at home I fooled around a bit and found that if I ran the blower really high I could get the smokestack temperature to about 550 degrees.  (I knew I could make the oil burner out perform  that  propane burner.) I couldn't get a good reading on the flue gases but did get a few indications over 650 degrees.   If the flues gases are really that hot, a smoke box superheater should work great.

Another piece of data ..... Dick McCloy and Dan Staron were watching one day when I was fooling with the blower to get rid of the smoke.   They noted how the fire seemed very sensitive to the amount of air for combustion.   I suggested that if I opened the firebox door, more air would be supplied and the smoke would go away.  The opposite happened, when I opened the door there was more smoke and I needed more blower to get rid of the smoke.  It was then obvious that air entering the firebox door wasn't used in combustion.

So, thinking about all the data I wondered how to get the optimum fire.    I now knew that air entering the fire box door and likely air entering the back of the fire pan wouldn't help combustion.   I had earlier cut a hole near the rear of the fire pan and installed a damper over the hole.  Experimentation showed that the best fire was with that damper closed but it didn't seal perfectly so I should probably seal that hole.  I concluded that the air  entering around the nozzle was the air used in combustion.   It then seemed obvious that I could get more combustion air either by increasing the blower or making a bigger opening around the nozzle.  Then I recalled the oil burner on the Heisler at Mt Rainier Scenic Railroad.  That fire pan is sealed except for a louvered opening around the nozzle; see MRSR 91 Burner.

The next step was obvious, make a new fire pan that was sealed except for an opening around the nozzle.  Make the opening around the nozzle bigger than before ( about X 3) and put  a damper on that opening. 

The next photo shows a side view of the new fire pan.  It's a bit crude --- this was an experiment.  I figured to learn a bit more and then make a nice final design.  The pan extends into the fire box about 3/4"  on the sides and back.   The bottom of the pan slopes up front to back.  The front extends down about 1.75"

All available area in the back is open except for about 0.5" on the sides and the nozzle mount.  The damper is a rotating flap  show in the open position above and closed position below.

The photo above shows the inside of the pan.  That is refractory cement used to insulate the pan.  The spark plug is inside a cylinder welded to the bottom of the pan.  I didn't have the correct tap for the plug so I made it a sliding fit and have a retaining plate on the bottom to hold it in place.

So how did it work?   Great!

When  the damper is closed, a very low fire can be maintained with very low blower settings.  As the fuel feed is increased, the damper can be adjusted in conjunction with the blower to get a stable burn with no smoke. Too much air cases the flame to be unstable --- too little gives smoke.  I was able to get at least 25% more fuel feed than before at a lower blower setting and no smoke.  

There is a  maximum fire than can be maintained.   What happens as the fuel, air feed (damper) and blower are  increased,  there is a point where the fire starts to shoot out the small observation hole in the fire box door and also out around the nozzle.   The flames shooting out any opening reminded me of flames shooting out around a crucible at  the aluminum foundry I once worked.    Small flames shooting out probably won't cause any problem except if the engineer reaches down to the blow down valve handle with an ungloved hand.  In normal operation a much lower burner setting would be used.

The fist test at the track went great.  I was able to pull a much heavier load up grades at a greater speed than before while maintaining steam pressure. (The only thing that marred the day was that I had some dirty fuel that plugged the filter a couple times.)

The burner now relights with the spark plug much easier than before.  Sometimes I have to make it a little rich to get it to start.

I of course was taking smokestack temperatures along the way and found the temperature was much lower than before.   About 400 degrees under most conditions and a maximum of about 500 degrees after running the burner with flames shooting out everywhere.   This means that there is more heat transfer to the water( mainly in the firebox) and less heat up the smokestack.  It also means that the temperature differential in the smoke box is probably too small to make a good smoke box superheater just like Ken Schroeder already told me.

I now remember the first time I fired up the Shay Dan Staron observed that too much heat seemed to be going up the stack.   It's now obvious that he was right.  The object should not be a hot stack, it should be the hottest  firebox  possible with the coolest stack possible.   There's a meet in a few weeks where there'll probably be a half dozen coal fired steamers that operate very well.  I'm anxious to take a few smokestack temperature measurements for comparison.    

The fire pan worked so well that I probably won't make another that has a better appearance. One thing remaining is to make a control in the cab for the damper.   I now leave it set about 75% open which is good for most operation.  If I want to turn the burner way down when sitting I can reach down and close the damper all the way.   The control will be probably similar the one used on MRSR 91.  

Update 10/23/2006:  This fire pan was modified by removing the variable damper and installed a mixing chamber; see New Burner.

Shay Project
NLW Home