May 12, 1916. THE COLLIERY GUARDIAN. 896 __________________________________________________________________________________________________________ ______________________________ When this oxide is no longer capable of taking up any further sulphur, a ready market is found for it, the present value of oxide containing 50 per cent, of sulphur being £2 per ton, and the revenue thus obtained pays for the oxide and the cost of labour on the purifiers. Composition of the Gas. After the gas has passed through the purifiers, it is taken to the engines. Its average composition is :— C02, 3-6 per cent.; CnHn, 2-6; CO, 7-6; 0, 0-3; H, 50’2; CH4, 30-1; N, 5-6 per cent. The calorific value of this gas varies between 500 and 550 British thermal units, the average value being 520 British thermal units. The total sulphur left after passing the purifiers is less than 50 grains per 100 cu. ft.; and the purifiers absolutely eliminate the last traces of any tar left in the gas after passing the benzole scrubbers, so that trouble from valve sticking is abso- lutely unknown, and not the slightest trace of tar has ever been found in the engine. The current from the generators is delivered to a 13-panel switchboard, consisting of one voltage regu- lating pane], three generator panels, one summarising panel, and seven outgoing feeder panels. A testing panel is also included in the power house equipment, and suitable means for testing motors .up to full load is provided by means of a “ Walker ” air brake* dynamo- meter. Load. The motors connected to. the mains aggregate approxi- mately 1,700 horse-power, and operate the whole of the coke oven machinery, the fans, haulages, belts, shakers, fitting shops, saw mills, etc. The lighting load con- nected averages about 70 kw. The daily load on the station reaches peaks of 1,050 kw., and the average load during the 24 hours would approxi- mate 580 kw. The station is running continuously, seven days per week. Methods of Running. During the initial running of the plant some trouble (overheating) was experienced, due to the diffi- culty of obtaining satisfactory mix- ing of the very rich gas and air. The method of dealing with this trouble was as follows :—The main gas pipe to each engine was reduced for a distance of about 10 ft. to an internal diameter of 2Jin., and on reaching the engine was led into the mixing chamber for about 6 in. Entering the mixing chamber at right angles to this pipe, and about Sin. above its end, was an auxiliary air supply, controlled by a dia- phragm governor. The gas ad- mitted to the mixing chamber was then diluted here by a certain, quantity of air. A fur- ther air supply, controlled by a hand lever from the driving platform, was led direct into, the chamber of the governor valve, where the diluted gas and air met. After passing through the chamber of the governor valve, the mixture was given a rotatory motion by means of a set of vanes, and was also very thoroughly mixed by being passed through a set of perforated plates. With this device no- overheating or pre-ignition is noticed, and the engines can be run up to and above their rated capacity without trouble. The exhaust gases are analysed at stated times or as the need arises, and the percentage of carbon monoxide is noted. The result aimed at in the analysis of the exhaust gases is to obtain an excess of air, consistent with the engine giving its full power. Gas Consumption. The curve for the gas consumption shows that during March 1914 a maximum of 15 million cu. ft. was used during the month. By regular analyses of the exhaust gases, combined with instructions to the drivers to drive by the pressure gauge, the consumption was, however, reduced to 12| million cu. ft. at the end of May. During the next six months the consumption was again reduced, the load remaining approximately at the same value. This goes to prove the important part which exhaust gas analysis plays in the successful and economical running of these engines. The average of good working results shows :—Carbon dioxide, 9-8 per cent.; oxygen, 2-2 per cent.; carbon monoxide, nil.' The present consumption is approximately 39 cu. ft. per kw. hour. The wearing quality of the engines is particularly good, and the repairs needed have been very small. The total cost of repairs, including wages and spare parts used, for 12 months is 0-037d. per unit generated, which cannot be called excessive. The most tried part of a gas engine of the type under discussion is the exhaust valve, and the material which has been found to be the best for the work is nickel steel. Cast iron valves, though they do not burn on the face, arc apt to break at the neck or collar, and a nickel steel stem, with a cast iron head attached usually gives trouble by distortion and through causing pre-ignition. The valves in the plant referred to have been in operation for over two years without' a single failure through breakage. With regard to the ignition, the low-tension magneto and coil transformer are most reliable, and no failure whatever of these parts has occurred. Sparking plugs, on the whole, appear to give very little trouble, the most frequent cause of failure of a plug being the widening of the gap, due to the burning of the points, and (less frequently) short-circuiting due to fouling by oil. Starting Difficulties. With regard to the difficulty of starting a gas engine, the following is the author’s experience. As previously mentioned, duplicate ignition systems are fitted to allow for easy starting, but not once in