December 4, 1914. THE COLLIERY GUARDIAN. 1171 small a current as to render the sparking at contact so mild that it will not ignite gas-air mixtures and also can reduce the sparking of the relay contact which closes the bell circuit to an inoperative condition as regards explosive mixture ignition, then, by the use of this system in conjunction with a bell fitted with a condenser, we have a safe system. The spark made by the relay in making contact for the bell circuit is entirely removed by putting a condenser across the two points of stud and contact, just as in the make and break of the bell. A system such as above suggested was then connected up. Each part was tried in turn. The bell, fitted with a condenser, was put in series with a variable battery and 10 ohm relay contact. The 10 ohm relay was for the time being operated by two cells and an ordinary bell push. The first experiment was made by placing the bell (condenser-fitted) inside the chamber and subjecting it to an explosive mixture of 15*7 per cent, oven gas and air. From one to ten cells were put into the bell circuit in turn and the bell rung by the relay. No explosion occurred, and the bell was therefore considered safe. The relay itself was then placed in the chamber, an explosive mixture as before admitted, and the relay actuated by an external pu^h to ring the bell (also external) with from one to ten cells in the bell circuit in turn. Though a distinct spark was visible at the platinum contact of the relay, no explosion occurred. A condenser was then put across the spark gap of the relay—the spark disappeared and no explosion occurred. Finally, the two “ knocking wires ” were arranged as before in the chamber (no file) so as to complete the relay circuit containing a variable battery of one to ten cells and the relay. In the bell circuit, which was external to the chamber, were the bell, relay contact, and five-cell battery. Experiment: Relay, bell and batteries, outside ch tmber. “ Knocking wires 33 inside chamber, which close relay circuit when in contact. i No. of cells Relay. ^xture in relay Result, mixture. circuit. 10-ohm relay... 15’7 oven gas-air ... 1 ... No explosion. 10- 9 1U 33 33 “* • • _ _ »3. 10- ,, ,, ... ,, ... 3 ... Explosion. It is evident, therefore, that with a relay the same danger of ignition arises as with the bell, for with three cells in the relay circuit ignition occurred. One saving clause arises, however, and that is that the 10-ohm relay could be operated from anywhere in a pit of normal size by two cells only. High Resistance. A relay was then obtained with ten times the resistance—namely, a standard gastight G.E.C. 100-ohm relay—and the above experiments repeated. Experiment : Bell, reliy and batteries, outside chamber. “Knocking wres33 closing relay circuit inside chamber. Relay. Explosive mixture. No. of cells in relay Result, circuit. 100-ohm relay. 15’7 coke-oven gas air... 1 ... No explosion. 100 ,, ,, ... 2 ... 33 100 33 »> ... 3 ... 33 100 33 33 ... 4 ... 33 100 ... 5 ... Explosion. 100 33 33 ... 6 ... The additional resistance, therefore, becomes a saving factor. No 100-ohm relay requires more than two Cells to operate it, as, averaging 13-ohm resistance to the mile of wire, the relay was found to be operative with two cells even when 100-ohms were placed as resistance in the relay circuit. We have thus a satisfactory and operative method for seven miles of wire. The relay is safe in a coke-oven gas-air mixture up to five cells, so that, allowing for all extraneous resistance, we have an absolutely safe margin for working even with four cells, as an oven gas-air mixture was only fired by six cells in the circuit, and a methane- air mixture would not (as our experience in the preceding experiments show) be fired when the relay was operated by six cells. The whole system, with the relay circuit containing four cells and closed by a file Contact on the “ knocking wires,” the bell (condenser fitted), and five cells in the bell-circuit, and the relay itself were placed in the chamber in a 15’7 per cent, oven-gas and air mixture. Although hundreds of contacts were made, no explosion occurred. The system is shown in diagram below. KN4QKIN* WIRES' CONttNSfA AC ROtS SPAA* GAP Fig. 2.—System Employing Relay, Condensers, and Shunt. There is with this method, as stated above, an abso- lutely safe margin of three cells over and above what is required for working—that is to say, that an explosion can only be brought about when six cells are used to operate the relay; but, as there is no necessity to use more than three cells, there is a margin of three cells on the side of safety. Use of a Shunt. To increase this margin, the investigation was taken a little further, using the suggestion that was made to the experimenters of introducing a method of limiting the inductance of the coils of the relay by introducing a shunt across the relay circuit. A resistance box, con- taining resistances of 3,000, 2,000, 1,000 ohms, was placed as a shunt across the relay circuit. On making contact with 2,000 ohms in the shunt there was an immediate diminution in the size and appearance of the spark at the contact of the “ knocking wires ” when using from one to 10 cells to operate it. One might actually say that with from one to five cwlls there was no visible spark at all, and with an increasing number of cells, though the spark was visible (and distinctly so when 10 cells were in operation), its character had altered from that when the shunt was absent, being a pure blue without any red colour whatever. The investigations were then transferred to the chamber. The relay was connected vid the galvanised iron “ knocking wires to a variable battery of Leclanche cells by which cell after cell could (as before) be put in the circuit. The shunt contained the resistance box and was altered in resistance as shown in the following list of experiments:— 100 Ohm Relay. No. of Current; Voltage ceils in circuit. Shunt resis- tance. Ohms. Leclanche 3,0"0... 10 ... 0T30 . 2,000... 10 ... 0T35 . 1,110... 10 ... 0 140 1,100... 10 ... 0T39 . 1,050... 10 ... 0T39 . 1,010.. 10 ... O’lil . 2,000 .. 10 ... 0’135 . We have, therefore, at at “ knock-“ knock- Mixture. Result, ing ing wires/* wires/* ... 14’75 ...15’7 oven air...No explosion ... 14’50 ...15’7 gas air ... ,, ... 14’50 ... ... 14’50 . . ... 14’50 ... ... 14’50 ... ... 14’50 ...Methaneair.. ,, >, by introducing a shunt into the relay ciicuit, increased the safety of the system to a point which does not require improving on the plea of safety, for there is no necessity to make use of more than three Leclanche cells, as previously stated, to operate the relay, even in the largest area over which a signalling apparatus is required underground. With a shunt across the relay terminals the relay is safe with 10 cells in circuit in an oven-gas mixture, and, therefore, with a much wider margin in methane air mixtures. Experiments have not been carried beyond a 10-ceU circuit, as it was thought that a further battery power would never be employed for signalling. In order to bring the result into line with colliery practice, a contact with file and needle was made, in the chamber operating the relay, with the shunt of 2,000 ohms and 10 cells in circuit. Hundreds of contacts were made and no explosion resulted. The connections for the additional safeguard of a shunt to the relay are shown in fig. 3. KNOCKINC wires’ Fig. 3.—System Employing Condensers and Shunt. There is little or no doubt that the case for some protection of signalling systems has been proved by each investigator, chiefly to be relied on being the investigations of Dr. Wheeler. In an approved apparatus (for it is modelled entirely on that used by the Explosions in Mines Committee) we have shown that the system illustrated above is safe even in oven-gas and air mixtures, and we can therefore recommend its use as an entirely safe system in the most gassy pits that can be selected. By arranging the relay so that gravity assists the break,” the system is one which will give little or no trouble, and it has the further advantage of being “fool-proof,” a qualification which is not an attribute of some of the sparkless systems now being offered; and, furthermore, by its very construction it means an enhanced life to the bell used. The author wishes to tender his thanks to the colliery electrician for the use of apparatus and for the sugges- tion of using a relay, which, it should be stated, had previously been the subject of an article by him on a suggested system of electric signalling appearing in a technical journal, and at that time with the object of diminishing the danger arising from sparks at break contacts. Immingham Coal Exports.—Returns for the week ending November 27, show that the coal exported from Imming- ham totalled 9,199 tons foreign, against 56,458 tons foreign and 2,860 coastwise during the corresponding week last year. Shipments -were as follows :—Foreign : To Aalborg, 1,373; Dieppe, 1,524; Gefle, 4,510; Landscrona, 1,393; and Tre- port, 399 tons. POWER PLANT TESTING.* By W. M. Selvey. In the early “ electric light ” days the largest item in the cost of power was the “interest and depreciation” in connection with the capital cost of the plant installed. The experience that the more successful undertakings are now passing through is of another kind. Time has shown that owing to the rapid advances in thermo- dynamic efficiency of power plant an item, which is now termed obsolescence, was omitted from the old balance sheet. It is the author’s opinion that unless provision is definitely made for replacing obsolete plant out of revenue, no longer than eight to 10 years should be allowed—if not on paper, at least in the mind of the buyer—for complete obsolescence of plant purchased at the present time and rated at less than 5,000 kw. in the case of turbo-alternators, or 30,000 lb. per hour for boilers. The author thinks that within the next 10 years the past, and to a large extent the present, significance of load factor will have disappeared. Further improve- ments, which will subsequently be obtained when the load factor is 50 per cent., will not be a vital matter in the cost of generating electrical energy. In the author’s opinion a strong case can be made out to justify very considerable care and expense in the testing of plant. The accuracy of the test should be worthy of the aims of the manufacturer. It is proposed to consider the various plant installed in a power station, noting their efficiencies and the methods avail- able for establishing the latter, and taking them in the following order:—(1) Boilers and economisers; (2) turbines and alternators; (3) condensers and air pumps; (4) auxiliaries—cooling towers, fans, pumps, &c. Boilers and Economisers. The efficiency of boilers has received much attention of late years, owing to the energetic advertisements of makers of CO2 apparatus. The use of the term CO2, meaning thereby the percentage of carbon dioxide in the flue gases, has become very common. It is a convenient way of denoting the amount of excess air in these flue gases over and above that which can take any part in the combustion of the fuel. In figs. 1 to 3 the efficiency obtainable is expressed in terms of the CO2 and the final temperature. Fig. 1 is ioo Efficiency of boiler Jdeal 8o 6o jo% m terms of excess air and temperature Dew points .__________________________________________ -p---- - ---- j|Tuei—-5% 160 26o 360 46O 56Q 660 Flue temperature in degrees Fahr (ACmosphcric CempenzCure » NOTES’.- Fan power omitted • , Staff f Slack coai c.v 12,210 B.Th.U,s(hi$he ) —Carbon....62-7%—------------—I-------— Hydrogen...-4-1% Moisture....8-6% Ash........10-4% Boiler radiation and unburnt|fuel—5% >60 Fig. 1. for a Staffordshire slack coal, and figs. 2 and 3 are for a Northumberland steam coal. The ideal percentage of CO2 varies with the ratio of hydrogen to carbon in the coal, but in most English coals the variation is not more than will cause the ideal amount of CO2 to vary from 18 per cent to 19 per cent. The method of collecting the sample has received much thought, and the author thinks that the arrange- ment shown in fig. 4 will be found to give reliable results. A considerable stream of gases is drawn through a sampling pipe, perforated with many fine holes right across the flue, together with aspirators capable of drawing off a steady fraction of this main stream. A little Roots blower, or what is still simpler, a small steam ejector, is employed for this purpose. The diagrams indicate that the other main measure- ment which determines the efficiency of the boiler is that of the final temperature. These two measurements having been established, there are only two important unknowns, namely, radiation and unburnt fuel; these the author has studied in considerable detail. As regards radiation, he is of opinion that the loss is no more than 3 per cent, in a 20,000 lb. boiler well lagged, and that it may be taken as 2 per cent, in a 30,0001b. boiler and larger sizes. Excessive radiation is always easy to locate, and not difficult to remedy. The loss by unburnt fuel can certainly be reduced to 10 or 15 per cent, of the ashes, and may therefore be taken as 2 per cent. For these reasons, in drawing figs. 1 to 3, 5 per cent, has been allowed for these two losses. It may here be not inappropriate to give the author’s opinion as to what constitutes good practice, which depends very much on the class of coal and the furnace construction. He has so far not seen a stoker and furnace construction that does not suffer either in efficiency or upkeep if the coal is either too bad or too good. If the boiler is designed for a good coal—and by this should be understood a coal giving anything above 12,800 B.Th.U.’s as fired, or below 8 per cent, of ash—it cannot maintain combustion at all with some coals. If, on the other hand, the boiler is designed for a bad coal (11,000 B.Th.U.’s, or less as fired, or above 16 per cent, of ash) the temperatures attained when working at its best, and with good coal, are such that no brickwork will stand. The best working can only be got by keeping the quality as uniform as possible, and this is * From a paper read before the Institution of Electrical Engineers.