The Crown Sheet Boiler

My starting point was to get a feel for what I had purchased.
There is an old Wall Street adage that goes like this, ‘Buy on
hunch, sell on facts.’ My own first reaction upon getting down
to details was that perhaps these really were intended by their
designer as hot water heaters and not steam boilers. Well, that
could or could not be academic at this point. The facts are that
here is a fired pressure vessel built to code requirements that is
capable of a 300 psig hydro test so how are we going to apply the
necessary boiler fixtures? Let’s start with trying to rate the
unit.

Although the total surface in the flue gas pass is something a
bit over ten square feet the effective heating surface is about 9.3
ft squared depending upon how one derates that portion of the
combustion chamber in the shadow of the burner. Of this, 25% is in
the radiant section (fire box) and 75% in the convection (tubes)
section. So, let us call it 9 ft squared effective heat transfer
surface.

My first thought was to refer to the famous Coatsville tests of
1912. Here Dr. Goss ran tests on a special design of fire tube
locomotive boiler in which the steam produced in the radiant
section or firebox was collected separately from that produced in
the convection section. He found that a coal fired boiler of that
type would produce 55 pounds of steam per hour per square foot of
heating surface. And, the figure was 10 pounds for the convection
section. In our case that would be 190! That’s way out of line.
His test boiler had 3200 ft squared and his results simply will not
scale down by a factor of 320:1.

So, let us use that old tried and true but now almost forgotten
term of 10 ft squared of heating surface equals one boiler
horsepower which equals 33,470 Btu/hour absorbed. Maybe that will
not be too far a field since I’m planning on supplying a one
horsepower steam engine.

Going that route would give us an output figure of 30,128
Btu/hour. Advertising for these units indicated a figure of 131,000
Btu/hour which seems out of line but don’t let that bother us
for we do have the dimensions and can establish the expected
performance on our own.

40,000 Btu ‘pancake’ burner with its original
pilot/thermocouple assembly. 10′ dimension later reduced to
8′ for better flame distribution.

I ran a hydro test on my unit at 200 psig. I believe that one
purchaser ran a 300 psig test but, here again, that is for another
story. Since this is a fire tube unit then the safety valve could
be set at 133 pounds. My plans were for a 125 pound setting and
operation at 100 pounds pressure. Since there are no super heater
or economiser surfaces, it’s simply a saturated steam
operation. As 100 psig steam contains 1189 Btu/pound and with 72
degrees feed water the heat absorption rate is 1149 Btu/#.

At 100 psig operation the saturated steam temperature will be
338 degrees F. That means that the flue gas exit temperature is up
around 500 degrees F and the boiler efficiency can hardly be better
than 70%. Dividing our 30,128 pick-up rate by the 1149 Btu/# of
steam says that we should be able to produce 26 pounds per hour of
steam. But to do that we will have to fire at the rate of 43,000
Btu/hour to make that 26 pounds of steam.

The combustion chamber measures 11.5′ diameter by 17.5′
long or 1.08 cubic feet. This gives us a furnace heat release of
close to 40,000 Btu/ft cubed at 100% rating. Now that is getting up
quite high for even gas firing. Full scale boiler furnaces burning
natural gas don’t run much over 30,000. So, we could be a bit
tight on furnace volume and we will need a good burner for either
gas or light fuel oil. Furnace volume may turn out to be the
limiting factor rather than heating surface.

With this as background I began to look around for a burner that
was capable of something like 40,000 Btu-/hour. Although I would
dearly loved to have had a coal fired boiler, practical
considerations dictated that it be LP gas fired.

Suburban builds a line of LP gas furnaces for the recreational
vehicle trade. They range from 22,000 to 30,000 Btu/hour heating
rate. These burners were designed for a modular combustion chamber
but which would fit the boiler’s 11.5′ diameter furnace
very well. A trip to the local RV repair shop specializing in
Suburban furnace repair located a wrecked 30,000 Btu unit… just
what the doctor ordered. The burner along with the safety controls
and pilot light assembly were removed. From there it was a simple
matter to mount these items on my unit and then borrow the 20# gas
bottle from the backyard grill.

There are those that, no doubt, will try to use the flat pancake
burner that is popular in domestic hot water heaters. The axis is
all wrong, in my opinion, The flame is concentrated in the center
of the chamber instead of throughout the length of the furnace. I
tried one at 40,000 Btu rate but really prefer the Suburban from a
design point of view though it is short on capacity.

The next project in the burner department will be to design and
install a steam atomizing oil burner. Since that is like holding
one’s self up by the boot straps, how does one handle a cold
start? One way will be to omit the LP gas regulator and use bottle
pressure (+60 psig) for atomizing medium until steam pressure is
available. That is another subject for another time.

One of the criticisms of these units is that to keep the 12
return tubes covered with water requires running a very high level.
True, but not worrisome. Look at it this way. Those tubes account
for only a small portion of the total steam production. Even if the
top row is uncovered there won’t be too much loss in capacity.
If this were a 20,000 pound per hour boiler of this design I’d
be concerned about thermal stresses. Their small size lends itself
to an inherently rugged unit.

But, that does bring up the subject of carry-over even without
priming conditions.

There are two 1′ outlets on the back of the boiler. One is
merely a pipe coupling welded into the back plate. The other is a
1′ pipe running to nearly the front of the steam release space.
At first, I thought that it was dry. pipe, but no, it’s just a
pipe open at one end. It can be made into what, in effect, is a dry
pipe by bushing down to a ‘ outlet to form a dam and then
stuffing the pipe with those non-soaped copper mesh scrubbing pads
used in the kitchen. One trade name is Chore Boy. These can be
strung on a piece of rod or heavy wire so that they may some day be
retrieved and also so that they will not find their way into the
boiler cavity. These make very effective entrainment eliminators.
Water mist impinges on the copper strands and coalesces into drops
of water that will flow back into the boiler.

I used the other opening for the safety valve. In my case I
bushed down to a ‘ street elbow and bushed it down to take a
‘ Kunkle safety valve set at 125 psig.

Speaking of water level. Trying to install a gauge glass is an
exercise in plumbing ingenuity. The two ‘ openings in the front
plate are not in line one above the other. There is nothing wrong
with a sloping glass but it would interfere with the chimney and
removing the front smoke box. After a bit of experimenting, I found
that a ‘ street elbow then ‘ by ‘ bushing, close
nipple, union and ‘ street elbow did the alignment job. A
nipple and a reducing coupling took care of the bottom
connection.

The feed water system is my ‘pride and joy’ and
something akin to a plumber’s nightmare. The original planearly
plans are always subject to change was to feed with a direct
connected boiler feed pump on the ‘main engine’ and with an
injector as back-up. For this and with a bit of studying FOR SALE
columns yielded a US size 00 injector. A ‘ gem! It was in its
original wooden box along with installation and operating
instructions to say nothing of a post card for the Plant Engineer
to send in for a free gift. The post card had a 1c stamp! Before my
time.

As it turns out, I have disconnected the pump’s connecting
rod after some test runs and now operate with the injector only.
For one thing, the water capacity in these units is enough that it
is not an onerous job to operate with the injector. Besides, its
like the VW bug’s gear shift it makes you feel wanted.

With all of the figuring and plumbing completed there came the
time for testing to really determine where things stood. The unit
was mounted on a test bed fabricated from 2′ by 6’s along
with a Stuart 5A (2′ by 2′) engine acquired in another
search effort. The test bed was fitted with a set of heavy duty
casters so that it could be trundled through the shop door to the
outside for firing.

From a cold start with the pancake burner at full firing rate it
takes about a half hour for steam to show with operating pressure
reached in another 30 minutes. At first I had the engine exhaust
into the stack for that fine sound of a steam engine at work. But,
unfortunately, the added draft doesn’t help the gas fired
system. In fact, it can pull the gas pilot away from the
thermocouple and shut the gas off.

In order to get some operating data on the system it was
necessary to run condensing just to recover the water to calculate
a steaming rate on not just the engine but to prove what the boiler
could produce. Now, since this is a story of a boiler and not of a
steam engine, the discussion will be limited to that part that
pertains to the boiler.

The engine was loaded with a prony brake. The exhaust steam was
condensed in a coil of copper tubing immersed in the copper wash
boiler we use to boil lobsters. A garden hose provided the cooling
water. The condensate line was fitted with a swing check valve and
the condensate then pumped out with the feed pump and collected in
a measuring can. The net result to all of this is that the boiler,
under the conditions that have been described and run at 100 psig,
will produce the estimated 26 pounds per hour.

The test series was extended to determine the maximum steaming
rate for the boiler. To do this, the pressure was taken up to 120
psig and maximum load applied to the engine. This turned out to be
40 lbs/hour steam rate averaged over several runs. But, the burner
did not have enough capacity to maintain this rate and the pressure
would begin to fall. The test run would be stopped at 80 psig. The
point here, with respect to the boiler, is that at this high rate
there was no carryover or other malfunction.

It can be concluded that the limitation for this unit lies in
getting the heat into the furnace and picked up by the setting.

The combustion chamber is full of flame with the 40,000 Btu
burner. It is not likely that a higher input rate could be handled
by the combustion chamber, but perhaps the heat transfer rate could
be improved.

The heat transfer from flame to heating surface is from the
radiant energy of the flame. With gas, this is from the glowing
carbon dioxide and the superheated steam (burning hydrogen) and it
is only at an efficiency of perhaps 30% of theoretical ‘black
body’ radiation. With oil or better yet, coal, the glowing
carbon particles which we recognize as the characteristically
yellow flame has a transfer efficiency much nearer ‘black
body’ transfer.

When we were racing the AFRICAN QUEEN against FRANCE II on Long
Island sound we were burning charcoal briquettes since in this area
of ‘the highly varnished set’ we couldn’t find any coal
or wood. It’s short blue flame made the job of keeping pressure
at 125 pounds almost impossible at full throttle. Just using the
bags that the charcoal came in by firing bag and all was a help
because of the luminosity of the flame. This leads me to believe
that using an oil burner on these boilers would let them produce
more steam than with LP gas.

These boilers are rugged little monsters capable of a lot of
model engineering fun. It will be interesting to learn of the
experiences of others.