108 Garfield Ave., Madison, NJ 07940
'The film is fiction, the boat is real.' I can still hear Jim Hendrick's voice as he answered the questions of reporters about the AFRICAN QUEEN then nudging the pier at the Norwalk Boat Show. It was about then, as I was trying to fix the wobble type bilge pump of the QUEEN, that I made the promise to myself to build a replica of that movie great, the AFRICAN QUEEN steam launch. I'd been writing about so many different steam powered projects that I figured that it was about time for me to do something myself. What is that old epigram, 'Those that can, do; those that can't, write'?. . . or something like that.
A great many home craft work shops have produced steam engines of a variety of sizes and form but it's the boiler that slows many down to a walk. The boiler would be my first operational step. That is when I came across an advertisement for a small boiler that seemed to just fit my needs. It seems that that same approach had appealed to many others for some 150 of those boilers had been delivered to steam fans and others at this writing. This is the story of what I did to make mine operational.
The item that I am referring to is that advertised under the banner of Crown Sheet Sales of Allentown, Pa. The specifics are a 17' wide by 24' high by 27' long boiler with an 11' diameter by 17' combustion chamber and twelve 1 5/8' flue tubes. I drove out to Allentown to take a look and when I first saw the one that I finally bought I termed it a Scotch marine boiler, though later, I was to revise this to be an internally fired, horizontal return tubular boiler. No matter, it only weighed 250 pounds and I could handle it on a tilt bed trailer, loading it by myself alone.
Antecedents of these units have interesting facets. Blue Ray built them originally for a duty not now clear, but financial problems forced them into bankruptcy, I am told, and the company's assets were taken over by Columbia Boiler of Lancaster, Pa. From there, Mr. Schaller of Allentown acquired the stock and began to market them under the nom de plume of Crown Sheet Sales. Crown sheet is a fitting name when one considers that he was, at one time, Treasurer of the Wanna maker and Southern Railroad, a steam powered tourist operation in the Pennsylvania Dutch country. Any old time steam railroader will tell you that you don't let the crown sheet run dry unless you have an appointment with the angels.
Out of the 150 or so units that I could determine had been shipped, maybe two thirds were to steam fans. The others went to very utilitarian projects from domestic hot water heaters to steam sources for commercial uses. Then there are something like, say, twenty that are intended for steam launches. Wouldn't it be interesting to learn how each of these have been fitted out for service? Particularly the one that went all the way to Fairbanks, Alaska; or the pair, one to Maine and one to Florida.
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.