producing 85 horsepower, with the same vacuum of 57.2 centimetres (22.5 inches) and with a higher pressure of 0.54 kilogramme or 77 pounds corresponding to an hourly flow of steam of 3,000 kilogrammes or 6,550 pounds. The accumulator this time is far more than adequate, and the lower pressure rises to 0 10 kilogramme or 1 pounds above atmospheric pressure, the upper pressure being determined (as before) by the valve exhausting into the atmosphere, as it did at each wind.

Fig. 12 indicates an analogous consumption of the turbine, the higher pressure being equivalent to 060 kilogramme per square centimetre or 85 pounds per square inch, vacuum 52 centimetres or 205 inches of mercury, and power produced, 65 electric horsepower. It will be noticed that here again the live-steam feedvalve was not called upon to enter into play, although the valve exhausting into the free air had been so adjusted as to open at a pressure only 0.25 kilogramme or 34 pounds higher than that of the atmosphere.* On the other hand, the valve admitting the live steam had been adjusted so as to let it in only at a pressure slightly less than that of the atmosphere, and this explains the lowering of the minima of the curve below zero.

In all these diagrams a peculiarly rapid rise of the pressure is noticeable at the time when the winding-engine begins to work. In concert with the colliery-engineers the writer, in planning the accumulator, had calculated that the winding-engine would eject steam during about minute (duration of movement of the cages); while in reality, as the graphic record proves, this exhaust only lasts over 15 to 18 seconds. Consequently, the speed at which the steam reached the accumulator was far greater than could have been foreseen, and the apparatus proving rather inadequate, it was found necessary to load the valve which exhausts into the free air, so as to ensure a rise or back-pressure of 0·40 kilogramme or 57 pounds per square inch above the atmospheric pressure, instead of the 0:25 or 0:30 kilogramme (3'6 to 4:3 pounds) which had been provided for. It would suffice, however, to add

*The exhaust-valve still allowed a great quantity of steam to escape into the atmosphere, even when the turbine was taking up nearly 5,000 kilogrammes or 11,000 pounds of steam per hour supplied solely by the accumulator, while the automatic steam-trap of the last named was ejecting 1,000 kilogrammes or 2,200 pounds per hour. Hence it may be inferred that the total consumption of an engine like that at Bruay is at normal speed much more than 6,000 kilogrammes or 13,200 pounds per hour.

!

to the apparatus a fourth unit, which should comprize about 10 tons of cast-iron, in order to reduce the back-pressure to the desired amount.

These same diagrams show also that, after each interruption of the inflow of steam into the accumulator, there is at first a very rapid and then a slower, diminution of pressure. Such variations in the rate of vaporization are evidently due to the fact that the whole mass of cast-iron and water contained in the apparatus does not immediately assume an equilibrium of temperature with the steam. The heat, of course, takes a certain time to penetrate fully into the body of the metal, and the delay thus involved produces a rise and then a fall in pressure, both far more rapid than if a temperature-equilibrium could be set up immediately. It is, however, easy to ascertain by calculation that the greater part of the metal and of the water contained in the basins plays in fact the part of a heat-accumulator.

Figs. 8 and 9 represent, on the whole, the normal working of the turbine, since this had been designed for a higher pressure of 0·90 kilogramme or 128 pounds absolute. If the whole system then produced only 200 instead of the 300 horsepower which it is capable of producing, this was entirely due to the unsatisfactory vacuum yielded by the condenser, which only reached 564 centimetres (22.2 inches) of mercury, instead of the 65 centimetres (25'6 inches) which had been reckoned with. For with 65 centi metres (25'6 inches) the expected 300 horsepower could certainly be obtained, as the factory experiments prove-whereof an account is appended at the end of this paper.

It may be mentioned also that the variations of the speed of the engine are of greater amplitude than will be admissible in another installation, as here they amount to about 3 per cent. This variation in speed results partly, it is true, from the irregu larity of the work which the turbine is called upon to perform. It is evidently due also to the variations of pressure in the accumulator above the turbine. As the pressure oscillates at each interval, that is, about every minute, between 1 and 1:40 kilogrammes or 14 and 20 pounds absolute, we are confronted with much the same result as if the work exacted from the turbine varied from minute to minute by about 30 per cent. As, however, this variation is not excessively rapid, since the rise and fall in pressure each take more than minute to come about, a fairly

sensitive governor has certainly all the time been needed to prevent the speed from varying by more than 2 or even 1 per cent. When the accumulator shall have been supplemented by a third section, and when the difference in pressure becomes only 0:30 instead of 0'40 kilogramme (or 4.3 and 57 pounds) the fluctuations in speed of the turbine will be considerably diminished.

It should be added that at No. 5 pit at Bruay, which (as previously stated) has been working for a comparatively short time, the winding-engine on the shaft where the writer's apparatus has been erected is not yet worked up to its normal capacity, since it winds only 700 or 800 tons instead of the 1,200 tons for which it was designed. It follows that, for the moment, the emission of steam is not altogether regular, and the automatic live-steam valve on the turbine calls more frequently and during comparatively longer periods upon the boilers for a supply than would be the case in an older-established colliery. But this state of things will quickly change for the better, pari passu with the development of the workings.

To sum up: despite the not very favourable conditions under which the writer's first steam-accumulator must needs work, it has already proved, and proves daily, that the practical and economical utilization of intermittently ejected steam is a problem that has now been solved.

At Bruay colliery, the improvement in the vacuum yielded by the condenser and an increase in the bulk of the heat-accumulator will in the very near future ensure the full advantage that may be expected from this first application. A fortiori will this hold good of new installations, for most collieries, from the standpoint which we are occupying, work under far more favourable conditions than No. 5 pit of the Bruay collieries. In steelworks, the circumstances should prove still more favourable.

The writer has thought that it might be of interest to summarize in the appended note a description of the low-pressure turbine and the experimental results obtained at the factory. These results were recorded partly by himself, and partly by Mr. Sauvage and Mr. Picou, in the presence of the engineers of Bruay colliery.

In conclusion, he wishes to express his thanks, first of all to the board of directors of the Bruay colliery and more especially to Mr. Soubeiran (who was confident as to the practical value of the writer's ideas); secondly, to the engineers of Bruay colliery, who took every care in erecting the various portions of this new installation.

APPENDIX.-RESULTS OF THE EXPERIMENTS CARRIED OUT IN APRIL, 1902, ON THE LOW-PRESSURE TURBINE FOR BRUAY COLLIERY, IN THE WORKSHOPS OF MESSRS. SAUTTER-HARLE & CIE.

The Bruay turbine, designed and constructed in the workshops of Messrs. Sautter-Harlé & Cie, is a multicellular turbine comprizing a series of wheels, 88 centimetres (34·65 inches) in diameter, adjusted upon one and the same horizontal shaft. These rotate, according to the arrangements that characterize the writer's system, between diaphragms, which are fixed within the cylinder of the engine, and on the periphery of which the distributors are set. The whole thus constitute a consecutive series of turbines with partial injection, which the steam traverses parallel and concentrically to the axis, passing alternately from a crown or ring fixed blades to a wheel or ring of moving blades, and from this last to the next following distributor.

The working being effected by impulsion, the expansion of the steam takes place only in the successive distributors, while the fluid acts on the moving blades by its vis viva. Therefore, each moving wheel rotates within a medium of uniform pressure which allows of considerable play between the blades and the adjacent fixed parts (3 to 5 millimetres, or 0·12 to 0.20 inch). The radial section of the blades and in particular the section of the various distributors increase the farther outward flow of steam and afford continuously wider passage to the fluid in proportion as its pressure and density diminish. The maintenance of a constant speed for variable loads is ensured by means of a ball-governor, which adjusts the higher pressure of the steam supplied to the engine to the power required. This governor is provided with a Denis compensator, which is designed to maintain the speed permanently at a fixed rate. Moreover, the speed may be modified at will between 1,500 and 2,000 revolutions a minute, thanks to a spring the tension of which can be varied by hand. Moreover, an automatic valve is adjusted, so as to admit into the turbine live steam coming direct from the boiler in case the supply of exhaust-steam proves inadequate.

Two continuous-current dynamos, supplying a triplex-wire 500 volts system, are mounted on the same shaft, directly coupled to that of the turbine. They are two-pole dynamos, but are also provided with two supplementary poles so that there may be commutation at the brushes without throwing these out of gear. The induction-coils, series-wound, can each furnish a current of 400 to 450 amperes at a tension of 250 volts.

Fig. 1 is a general view of the machinery, reproduced from a photograph, The experiments conducted with this turbine lasted in each case about 20 minutes, and allowed of 8 series of observations every 3 minutes. The average or mean results are embodied in Table I.

TABLE I.-EXPERIMENTS WITH THE LOW-PRESSURE TURBINE AND DYNAMO OF 300 HORSEPOWER FOR THE BRUAY COLLIERIES.

Revolu

Absolute Pressures.

Temperature
of Steam.

Flow of Steam
Measured.

Consumption of Steam per Horsepower-hour.

tions

per Minute.

Volts.

Am-
pères.

Kilogrs. Pounds. Kilogrs. Pounds.

1,337 4710 3600 169.5

230 0

0.845

12.02

0.150

2.13

132 269-6

20.50

45.2 9.95

21-90.485

1,412 475 0 364 0

19.85

43.8

0.501

دو

19:40 42.8

10:05 22-2 0.518

4,660 165 18:40 40.6

APRIL 5TH, 1902: EXPERIMENTS MADE BY MESSRS. SAUVAGE AND PICOU.