GAS ENGINES FOR SHIP PROPULSION.

By J. E. THORNYCROFT, Esq., Associate.

[Read at the Spring Meetings of the Forty-seventh Session of the Institution of Naval Architects, April 5, 1906; the Right Hon. the Earl of GLASGOW, G.C.M.G., LL.D., President, in the Chair.]

Ar the Spring Meetings of the Institution of Naval Architects in 1904, the use of gas engines was referred to in a paper read by the author on "Internal Combustion Engines for Marine Purposes."

At that time there was practically no information available with regard to them but just then Herr Emil Capitaine was carrying out experiments with his first boat, and a number of vessels have since been fitted which have shown very satisfactory results, and it is thought that a description of them, and particulars of the results obtained, will be of interest to the Institution, Sir William White having recently stated that he considers the gas engine will be bound to play a most important part in marine engineering.

Before giving details of the different vessels, it may be well to shortly describe the way a gas engine works in conjunction with a gas producer, and consider the possible advantages and some of the difficulties that have to be overcome.

The majority of small gas engines work on the "Otto" cycle, which has the advantage of requiring no special gas or air pumps, and this is the only type of engine that has been as yet tried for marine work.

The gas producer which is generally used for moderate powers works on what is known as the "suction" principle: that is to say, instead of the gas being generated by the combustion of fuel, by air and steam being forced through it under pressure, it is generated by the air being drawn through the producer, the whole of the apparatus working somewhat below atmospheric pressure. The advantage of this system is that neither a steam boiler working under pressure, nor a gas container for holding the gas is required, and the further very important advantage for marine work is that there is no danger from a leakage of gas from the producer; if leakage takes place, it is from the atmosphere into the producer itself.

There are so many combined plants consisting of a gas engine working in conjunction with a suction producer at work for stationary purposes, that it might appear only very small alterations would be required in these plants to adapt them for marine propulsion; but, in addition to the great difference between almost all marine and land plants, viz., that of the quick and frequent variations of load which the smaller types of marine plants have to undergo, there are very many features which must be changed to adapt the gas engines and producer for boat work. In fact, it will be found that the principle of operation will be about the only thing which can be identical with the land plant; the details of the apparatus all having to be specially worked out for the new conditions.

The producers for moderate sized powers are usually worked with anthracite coal, and where vessels are to be used on fixed routes, and the same class of fuel can always be obtained, anthracite or coke will be found to be the most suitable for moderate powers, it being too difficult to make a satisfactory small plant to work with bituminous coal.

For larger powers, and for vessels that have to take up fuel at different ports, the producer must be capable of dealing with coal, other than anthracite, which may be of a tarry or caking character.

The design of the producer capable of successfully dealing with bituminous or caking coal is a much more difficult problem than the anthracite producer. There are, of course, very many large land installations of pressure producers working with bituminous coal, but the apparatus for extracting the tar and other by-products, as at present arranged, is far too cumbrous for marine work, and the marine producer to deal with this class of coal must be designed on entirely different lines, being arranged to consume the tar which is evaporated off from the coal in the first stages of its heating. This has been arranged in some producers by taking the gas off at a comparatively hot part of the fire, instead of at near the top of the new coal, as is usually the case with the anthracite producer; but there are plants now at work dealing successfully with bituminous caking coal, which are fed by an undertype stoker, so that the new coal is forced into the hottest part of the fire, and the tar which is driven out is at once burned.

After the gas leaves the producer, it must be first cooled, and, as there will always be a certain amount of impurity, even when anthracite is used, it has to be thoroughly cleaned. This is usually done by passing the gas through a series of vessels where it is scrubbed by its passage through layers of coke over which water is running.

It is frequently arranged that the cooling of the gas is effected by passing it over a water-containing vessel, which acts as a generator for the steam required by the

producer; but there are disadvantages in this method where the producer is not working at a uniform load, as the steam will not be generated until the producer has been working some time, and will continue when the engine has been stopped, and when it is not required, and so cool the fire, causing the gas to be poor when the engine is again started, after the temporary stoppage has taken place.

The large space occupied by the ordinary coke scrubbers prohibits their use for marine work, and to meet this difficulty Herr Capitaine has arranged his plant to clean the gas after it has been cooled, by the introduction of a very fine spray of water, which mixes with the small particles of dust and other impurities in the gas, and forms a sort of fog; the gas in this stage is passed into a centrifugal apparatus, having a peripheral speed of 160 ft. a second, which throws out the moisture and impurities, leaving a clean dry gas to be drawn out by the engine.

The composition of this gas may be taken to be about as follows:

It will be realised that the size of the producer for a given power is comparatively `small when it is known that the area of the fire grate necessary is only 0.05 sq. ft. per horse-power, whereas the average for an ordinary natural-draught steam boiler, burning 15 lbs. coal per square foot grate area, would be 2 sq. ft. per horse-power.

The depth of the fire will vary considerably with different qualities of coal used, but will not be usually more than two to three times the diameter of the fire grate.

DESIGNS OF DIFFERENT TYPES OF PRODUCERS.

Fig. 1 (Plate XIV.) shows Messrs. Dowson's arrangement of suction producer designed for anthracite coal. The majority of the smaller suction producers will be found to closely resemble this type. A fire-brick lined steel casing is fitted with fire bars and a closed ashpit, and in the upper part of the casing a conical hopper is placed with the usual valves for opening and closing to admit fresh charges of fuel. The gases come off the fuel at the base of the conical hopper, and pass through a steam generating pipe or vessel, and then away to the cooler and scrubber. The producer is fitted with an outer casing in which the air is heated, and mixes with the steam on its way to the ashpit.

fuel.

Fig. 2 shows the Duff-Whitfield producer arranged for working with bituminous In this producer, the tarry products which are evaporated off from the new coal

are caused to pass through the hottest part of the fire, and so are consumed before the gas passes away to the cooler.

Fig. 3 shows the "Boutillier" producer arranged for working with bituminous fuel, in which the fuel is supplied by an undertype stoker to the hottest part of the fire, so that the tar is at once decomposed, and cannot pass away with the gas. The principle upon which this producer works seems to be one of the best adapted for marine work.

in

The "Jan" producer consists of a series of producers which are arranged to work sequence, the gases of one which has been newly charged passing through another which has been at work some time, so that there are always some of the producers supplying tarry gas, the tar being consumed by one of the other producers in the later stages of the series.

The heat losses in the producer itself are very small, as the whole plant can be so thoroughly lagged that practically there need only be heat passing away with the gas that is being generated. The loss from this cause may be very much reduced by using it to heat the air supplied to the producer.

The heat efficiency as given by Messrs. Dowson for their suction plant is 90 per cent. As the result of a great many tests, the late Mr. Bryan Donkin gave the average heat efficiency of steam boilers at 66.7 per cent.

To start up the Capitaine producer, a fire is lighted on the bars, and the producer is then filled with fuel, air being either drawn or blown through it by a hand or powerdriven fan. For a producer of about 50 to 100 H.P., twenty to thirty minutes should be sufficient to give good gas. This can be ascertained by lighting the gas which is being given off, and seeing that it continues to burn with the right sort of flame. The gas first produced will have very little hydrogen in it, and will not be very powerful, and requires the steam generator to come into operation. The steam which is added to the air entering the producer strengthens the gas by the water being decomposed, and gives the percentage of hydrogen which is shown in the analysis above referred to. The gas produced in this way is of a comparatively slow-burning nature, and may be described as poor gas, to distinguish it from town gas, and some of the other gases generated by pressure producers which have a larger proportion of hydrogen, such as water gas.

There is an advantage in this slow-burning gas, as it enables the engine to work with a high compression, thus giving a larger range of expansion, and consequently a high economy. The slow-burning nature also avoids the violent. shocks which occur when more powerful gases are used, and so a smooth-running

engine is obtained. The producer once being set to work will continue to generate gas in large or moderate quantities as the engine requires it and sucks it from the producer. The producer will probably be giving the best gas at somewhat near its full capacity, but will continue to give good gas down to a small proportion of full load. At very light loads it is difficult to keep the fire hot enough to produce good gas, and special means have to be adopted to keep the producer hot enough on these occasions.

Having now generally described the producer, the difficulties and advantages of adapting the gas engine itself to marine work will be considered.

The one great disadvantage of the internal combustion engine is the necessity of setting the engine in motion before it will run automatically. For powers less than 200 H.P., it is preferable to employ a reversing gear, keeping the engine always running in the same direction, or to use a reversing propeller. Compressed air is being employed for starting up large engines, and when once the engine is fitted in this way, the valve gear for running the engine in either direction does not amount to very much.

For moderate powers, a single acting engine with a trunk piston is found most convenient, as the piston does not require to be water-cooled, until one as much as about 2 ft. in diameter is employed. A single cylinder of 20 in. diameter and 2 ft. stroke, running at 120 revolutions per minute, will give about 100 H.P., taking the average working pressure at about 80 lbs., which is less than the figure often obtained.

The same sized cylinder, working as a double-acting cylinder, would, of course, double the horse-power; but, besides the additional valves, which must be an exact duplicate of those required for a single-acting engine, there is the complication of the water-cooled piston and rod. This at first sight will appear somewhat serious to the marine engineer, but when the very large number of engines of this design which are working on land are considered-the regularity with which they are run, and the little attention they require-it will be agreed that there does not seem any reason why they should not meet with success at sea.

For small plants of from 100 to 200 H.P., a small auxiliary internal combustion motor, driven by oil or some other fuel, is found most convenient for blowing up the producer, for starting it, and also for starting the main engine.

For larger powers fitted to vessels where steam capstans and steering gear are fitted, it is thought that the best plan will be to employ an auxiliary boiler, which can be heated by the gas when the whole plant is at work, and can be used independently to drive the various auxiliary and starting engines when the producer is not alight.