More Power
Petticoat - Nozzle - Fire Pan - Feed Water Heater
Nelson Riedel
Initial: 11/26/08 Last Revised: 12/03/2008


A couple really busy years have passed since the last major improvements to the Shay  ---- volunteer teaching trips  plus design of a new house have limited the activity with the Shay.   We moved into the new house at the end of August and the shop was operation in operation a couple weeks after.  (An update of the shop description is a good project for the cold winter.)

Meanwhile --- the construction at Mill Creek Central Railroad has proceeded at a brisk pace with the completion of the Mountain Division reverse loop in late spring 2008.  See the Track and Structures and 2008 Construction pages at   The elevation graph below shows the killer grades at MCC.

This fall I fired up the Shay and tried it up the big hill ---- from Barney Yard up Wilson Siding, through Varian Tunnel, across Tower meadow, over the trestle and up the final grade to the peak just beyond the gas well.  How did it go? Embarrassing!   The Shay had to go very slow ----- like 5 scale mph.   I couldn't add water on the grade or would loose pressure and would have to stop to rebuild it.  Had to stop at Tower siding to add water to the boiler and rebuild pressure.

The Shay has been able to make the grade from Daffodil siding to Barney yard with no problem.  I usually stop at the yard to add water to the tender so it was no problem if I got there with low water in the boiler and low steam pressure.  The big difference is that the grade to the gas well is much longer.  


Note that between Daffodil siding and Barney yard the track climbs about 15 feet in a distance of about 1300 feet.   By contrast, the climb from Barney yard to the high point just beyond the gas well is about 35 feet in a distance of about 2700 feet.   It's clear that the performance on the long grades is limited by the ability to generate steam.   This is not a problem on a  railroad with limited grades --- the Shay can easily pull ten or more cars on flat or nearly flat tracks.   

The simple solution is to just open the oil valve.   Unfortunately,  increasing the oil causes the fire to smoke and quickly soot the tubes --- little extra heat is generated.   Increasing the blower pressure helps some but the steam through the blower is lost --- it would better used to drive the pistons.    The exhaust seems to have little effect on the draft --- not what one would expect.  

The biggest problem seems to be insufficient draft.   The first step was to take a look at the smoke box to see if there was a way to improve the draft.

This shows the existing smoke box configuration.  The diagonal pipe is the blower.  The vertical fitting on the right is an elaborate plug --- a pressure relief valve was previously connected there to prevent high pressure from damaging the blower pressure gauge.    That pressure relief was removed several years ago.
This is a photo of the petticoat of Mount Rainier Scenic Railroad  91 - a late Heisler.  Note that there is an opening between the top of the petticoat and the bottom of the smokestack.  I couldn't find any photos of the prototype Shay petticoats.  However, I seem to recall that the shays have a similar opening at the top as well as one or more gaps part way down the pipe.
This shows how I modified the shay petticoat.   On the original design, the petticoat slides over the part of the smokestack that sticks down into the smoke box.   A piece of thick tubing was inserted between the smoke stack and the petticoat.   The tubing was turned so that the bottom slides into the petticoat and the top slides over the smokestack.  Large holes were then drilled near the top of the tube and through the part of the smoke stack that was inside the tube (the smokestack was removed for this machining). .  These holes were then enlarged on the mill.  The holes are a little over 5/8" high and more than 3/4" across.   Note that the bottom of the petticoat is about 1.5"   lower than before.
Update 12/3/2008

I received several emails with comments and input after posting this page.   Mike Green sent a couple photos of Shay s/n 3345 smoke box shown below.   This input  stimulated me to look at my reference material again for any photos or sketches of the Shay smoke box.  This time I remembered some reproductions of Lima Locomotive Works information.

The smoke box side view figures on the right are from  a reproduction of INSTRUCTION SHEET No. 9 - Firing and Drafting of Locomotives.  Fig No 1 is similar to the MRSR 91 Heisler smoke box shown previously.  That Heisler is oil fired.  Fig No 2 is for oil and wood fired engines.  I seem to recall seeing a two or three stage petticoat  (two or three inverted funnels) laying on the floor in the Cass Shops which was similar to Fig No 2.  All the locomotives at Cass are now coal fired.    Yes --- these data is in conflict with the two figures. 


This is from a Shay Repair Parts Book reprinted at Cass.  Toward the front of the book  the S/N of the Cass Shays are listed and the pages in the book that would apply to each locomotive.    This image was taken from page 20 which was listed as applicable to S/N 1503 --- Cass No 5.   This seems to match Fig No 1 above.    
This is one of the photos of the s/n 3345 smoke box that Mike Green sent.  There is a clear gap between the top of the petticoat pipe and the bottom of the smokestack.
This is the bottom of the s/n 3345 petticoat.  Note that the exhaust nozzle is close to the bottom of the smoke box similar to Fig No 2 above and similar to the photo of the MRSR91 smoke box.  Mike said these were very old photos.   Information at the  Shay Information Website indicates that s/n 3345 started life as a coal burner but was modified to be a wood burner during the recent restoration.  Since the photos are old and the smoke box looks pretty rough, they were probably taken before the restoration.

End 12/3/2008 update


The effect of the changes to the petticoat was very significant.   Previously, the blower pressure had to be set to 30 psi or greater to prevent significant smoking at the lowest fuel setting that would provide a stable fire.   After the change the burner would operate at below 15 psi blower and there seems to be a  greater range of useful oil feed and blower settings.  At the track I found the performance on the big hill was much better ---- it was no longer an embarrassment.   With these changes to the petticoat the exhaust provided some draft and the blower could be set to about 10 psi with open throttle.  

Dick McCloy suggested that we run a test with the atomizer and blower supplied from a portable air tank --- the Shay plumbing is such that air can be supplied to the atomizer and blower for startup.  We charged a ~10 gallon portable tank to ~ 100 psi and started up the hill through Wilson siding with the blower pressure set at 15 psi.   The tank was essentially empty before we got to the tunnel.   The message is clear ---  the blower really sucks the steam --- we should try to generate more draft from the exhaust.

The Shay went back to the shop for the next set of changes.  One change was to reduce the diameter of the exhaust nozzle in an attempt to create a greater draft from the exhaust.  The nozzle was 7/16" diameter in the original design.  I decided to reduce the opening to 5/16" diameter --- half the area. 

The sleeve at right was used to reduce the nozzle.



This shows the sleeve in place in the nozzle.  The sleeve is a force fit.

Later tests showed that the sleeve was effective ---- more data further down.

The next thing tried was a larger fire pan to see if the greater volume would permit a greater amount of fuel to be burned.  While the old  fire pan was off I snapped the photo on the right which shows the deflector plate that is installed in the fire box to force the heat to the back of the fire box  and then across the bottom of crown sheet before exiting out through the tubes.  Note the oil that has been thrown by the engine onto the outside of the firebox where it dripped down and flowed across the bottom of the fire box.  Shays are dirty!
The previous fire pan was sloped --- the front with the oil nozzle hung down about 1.5".  The pan sloped up front to back and was even with the bottom of the fire box at the rear.

The new fire pan as shown at right is rectangular ---- the hope is that the greater volume will allow more fuel to be consumed and more steam generated.  The fire pan is constructed of 2" channel and 1/8" plate.    The four channels are attached to the plate with 6-32 screws.  This pan is attached to the firebox with the same 4-40 screws used on the previous fire pan.  (If I were to do it again I'd use 6-32 screws here.) 

The two tubes that run through the fire pan are to pre heat the steam for the atomizer (3/16" tube on the left) and the oil (1/4" tube on the right).  These tubes are stainless steel. 


This is the view of the fire pan from the rear.  

I don't have a tap to match the spark plug threads.   I drilled a hole just large enough for all of the spark plug except the hex part for the wrench to slide though that rear channel.   The plug is held in place by the plate which is secured to the channel with a 6-32 screw.  

This new pan was tested at home before taking the Shay to the track.    The burner didn't operate well at all --- I couldn't keep it running.   Maybe the oil was vaporized and the atomizer didn't work well with oil vapor.  I removed the fuel tube and ran it under the fire plan.  The burner then functioned properly.

So, I learned that superheating the oil doesn't work!  No wonder there is no mention of heating the oil in the literature. 

One inch diameter vent holes were drilled toward the rear of each fire pan side for additional vents.  Plates were made to cover all or half of each hole so that the effectiveness of the vents could be evaluated.  The photo shows one of the vents half covered.  The welds on the bottom of the fire box present an uneven surface thus preventing a tight fit of the fire pan to the firebox.  The photo shows that fire is visible at the joint.     These gaps  prove some additional venting   

These modifications really improved the power.  The Shay pulled the 6 cars shown above with a husky engineer and brakeman up the grade to the gas well without stopping.   The speed was 5 to 10 scale mph on the steepest grades and 15 scale mph or greater on the milder grades.  I was able to add water as we went along and had a full boiler when we got to the top.    One thing was different ---  the axel pump couldn't quite keep up with the consumption ---- maybe I'm able to burn more fuel.   I used the steam powered pump to add some water when we went through Tower siding.   The smaller exhaust nozzle greatly increased the draft ---- I could both turn off the blower when the throttle was open  and also increase the amount of oil slightly.   The smaller nozzle also produces a more pleasing exhaust sound.  There was one problem ---- the fire blew off more often.   On the other hand, it seemed to be easier to relight with the igniter.   Just after this photo was taken, the  operation became more erratic--- the fire seemed to be too lean and completely opening the valve had little effect.  The fuel filter was blow out at the beginning of the day but it seemed to be partially plugged again.

It was decided to take the shay back home to exercise it on the test stand to try to make the burner less prone to blowing out.  (It may be that the filter was partially plugged and the fire was too lean.)   Dick McCloy had another suggestion --- attach the output pipe from the axel pump to the under side of the fire pan to make a feed water heater.                 

The first thing I did after the shay was back in the shop was to disconnect the fuel line at the atomizer, open the fuel valve and drain the near full tank of oil.  The flow was very limited.  After an hour the flow slowed to a drip and 3/4 of the fuel was still in the tank.  A shot of air back though the fuel line and filter got things going again.   Had to do that once more to get the tank empty.    The photo at right shows the problem --- a sludge at the bottom of the fuel tank which plugged the filter.  This happened once before several years ago.  It is my understanding that the sludge is caused by the growth of microbes in the water at the bottom of the tank.


So --- I leaned that when the fuel filter plugs up, microbes are probably growing in the fuel tank.  The fuel should be drained and the tank cleaned.  It's probably a good idea to drain the tank each fall and clean it.  That should fix the problem since it seems to take several years to get the growth started. ( I also learned that the odor from the diesel fuel I'm using can linger in the house for many hours and some women find that odor unpleasant.)

After cleaning everything and putting fresh fuel in the tank I fired it up on the test stand (bearings under the wheels).  

All the problems with the fire going out seemed to have gone away.    The fire would burn with essentially no smoke at 10 psi blower setting and a low fuel setting.  The fuel valve could be fully opened and there was very little smoke with the blower at about 30 psi.    The burner seems to be about 2:1 range from low to high.   The limiting factor for the amount of fuel seemed to be the fuel nozzle itself.   This is good news ---- I seem to be using the full capability of the nozzle.   It also tells me that the under construction Heisler which is a bit larger than the Shay should have two oil nozzles.    

The exhaust was able to provide sufficient draft and the blower could be shut off completely when the throttle was partially open.   Opening and closing the throttle did not cause the fire to go out except at the lowest fuel settings.     

The oil feed line was under the fire pan.  I tried pushing it up against the bottom of the fire pan.  The operation seemed to get erratic --- so I concluded that running the line below the fire pan with a little gap provided about all the oil heating that can be tolerated and still have reliable operation.

During this test the holes in the sides of the firebox were covered.  I uncovered one of the holes to see if the operation was different.     There was not a significant difference good or bad.    However, with low draft settings the fire would shoot out the side through those holes.  That is not so good so it was decided to keep those holes covered.

While writing this up I realized that it sounded almost too good to be true ---- especially the fact that I could completely open the fuel valve and not get much smoke.  I decided to test it again to make sure before making any other modifications.   I blew out the fuel filter to make sure it was really clean. I also topped off the fuel take so I'd have maximum fuel pressure.    

It was a cold day  --- 26 F at 10 AM when photo above was snapped.  This is just outside the shop door.    (Earlier in the day a bunch of ducks were swimming in the lake.   Don't understand why they didn't fly south for the winter.   Maybe I should take pity on them and invite them in for dinner --- they are great with a cherry sauce.)    

With the full oil tank I was able to make the fire too rich.   However, even when too rich it didn't seem to smoke as much as before the modifications.   With the very cold weather it was difficult to tell whether the exhaust was all steam or whether there was some smoke.   In any case, I can't think of anything else to do to  generate more heat from the burner.

I ran the axel pump a little during this test and was reminded of the cooling effect of adding water so that probably the best next improvement to work on.

When  I took everything apart to make the feed water modifications I also cleaned the tubes.  There was considerable soot.   I suspect that the soot accumulated on the previous outing when I was having trouble with the plugged filter and the burner going out a lot.  I also recall that I forget to  add the  RED DEVIL SOOT REMOVER the last time I filled the fuel can.   This is probably a good reminder to not get too cavalier with the oil mixture.  


Photo above shows the fire pan after the water feed lines have been routed through it.    The copper tube down the outside is the oil feed.     The smaller stainless tube only visible on the left is the atomizer steam feed --- it is unchanged from the previous photos.    The stainless tube across the top of the photo is for the feed water from the steam powered pump ---- this tube was previously unsuccessfully used for heating the oil feed.  The U shaped tub is for the feed water from the axel pump.  The atomizer steam tube is 3/16" OD and all the other tubes are 1/4" OD.   All the tubes are connected with compression fittings to facilitate removal for maintenance.

The modifications were tested on the test stand before making the trip to the track.    It was about 35 degrees F during the test.   I used an infrared thermometer to measure the following temperatures while the boiler pressure was about 110 psi.

  • Boiler above crown sheet - 305 degrees F

  • Smokestack base - 305 degrees F

  • Axel pump feed water  when engine was running at ~ 10 scale mph - 190 degrees F

  • Steam pump feed water -- moderate pump speed ~ 130 degrees F.     

The boiler and smokestack base temperatures are a little low for 110 psi pressure ---- ~340 degrees would be expected and indeed temperatures near 340 degrees have been measured on warm summer days.   Guess the cold temperatures had some impact on the measurements.  Previously, the top of the smoke box was quite a bit warmer than the boiler ---- the holes in the smoke stack at the top of the smoke box probably allow those hot gases to escape.   Don't know whether that is good or bad.

 The feed water heaters did work --especially the axel pump heater.   The tender had been filled in the shop so the water temperature was probably about 60 degrees.   The temperature of the output of the feed water heater was measured by shining the infrared thermometer on the water tube about 6" from the fire pan.  Based on the measurements, the cooling effect of the feed water from the axel pump is reduced by at least 50%.   The next step is a field test on the big hill at Mill Creek Central.    

Went back to Mill Creek Central for a final test run on the Sunday before thanksgiving.   It was a few degrees above freezing when I did the run.  I had one car attached with two passengers.  The performance was very good in that we made it to the peak just beyond the gas well with essentially full pressure and a boiler full of water.    We made good speed on the way up --- ~10 scale mph on the steepest grades and about twice that on the less challenging grades.  

I seemed to be able to burn more fuel without significant smoking.  Dick McCloy commented that these was much less of a trailing oil smoke stench.    The burner operation seemed very stable.  It only blew out a couple times, each time apparently due to a too lean of mixture ----- as the oil level in the tank goes down, the valve has to be opened more. The total run time was about an hour.   I ran with some blower -- about 10 psi even when the throttle was open some.   

The feed water heaters did work and contributed to the performance.

After the run I cleaned a few of the tubes and found a little soot.    This probably happened when I ran very rich for a while to see if it improved performance --- it didn't.  At the rate it accumulated I would probably not have to clean the tubes more than once every 8 hours of operation.   The next challenge will be to  understand the burner well enough to be able to set the oil to the maximum possible with no soot accumulating. 

Further tests must await warm weather --- probably next April.  


  1. The biggest improvement  came from the changes to the petticoat and the part of the smoke stack that hangs down into the smoke box. .  These changes doubled the efficiency of the blower.

  2. The next biggest improvement came from reducing the exhaust nozzle diameter which together with the changes listed in 1 creates sufficient draft so that the blower can be shut off when the throttle is open.  Over half the improvement came from changes 1 & 2.

  3. The larger fire pan together with changes listed above seem to allow more fuel to be burned.   I don't have a measure on this but suspect the improvement is in the 10% to 20% range.  

  4. Significant heating of the oil caused the burner to not operate properly --- the fire was erratic.  Heating the oil doesn't work!

  5. Heating the atomizer steam seemed to improve the burner operation ---- this is probably a less than 5% effect.

  6. The feed water heaters are a significant improvement, especially the heater for the axel pump feed water.    These changes probably account for on the order of 25% of the improvement.

I expect to incorporate all of these changes into the Heisler design. 

Shay Project
NLW Home