If we are to produce an average of one electrical horse-power per hour, we shall produce twenty-four horse-power hours during the twenty-four hours, half of which must be stored; viz., we must have storage battery capacity for twelve horse-power hours. Of course, the rate of discharge will have some bearing upon the cost of this storage battery capacity, but, taking an abstract case, we may assume, I think, without being unfair, that it will cost $35 per electrical horse-power hour capacity erected. We must, therefore, add to our former figures, $12 x $35 = $420, which is more for the batteries than we have allowed for all the rest of the plant put together, including the buildings. The total fixed investment becomes Interest on $726.80 at 4 per cent. Machinery, $181.80 at 72 per cent.. therefore $726.80 29.07 2.50 13.64 42.00 $87.21 Total fixed annual cost. Coal, 2.95 pounds for 8,760 hours, at $1.75 per ton.. Petty stores, attendance, etc.. Total annual cost per electrical horse-power, $22.61 $123.13 which would be an actual loss over that of irregular working of $123.13 $117.78 $5.35. It will cost with the storage battery thus used, $123.13 - $48.68 $74.45 more than the same machinery, working continuously on a steady load, could produce it for. If we assume coal to cost $3.50, instead of $1.75, the cost per electrical horse-power will be: Steady load without storage battery.. $52.31 148.40 145.74 We thus see that the storage battery, on account of its extreme cost, is entirely out of the question as an economical device when thus used. This does not condemn its use in central stations in other ways, however, for there are many cases in which its use. may contribute to economy, but these cases cannot be determined by any general rational formula. Each case must be determined for itself, and the result will depend primarily upon the shape of the load line, and secondarily upon the ability of the battery to rise above normal discharge rate economically. In the case of a station previously equipped, without storage, so that the units are already determined, the question of economy of introducing storage batteries will be determined in some cases entirely by the sizes of those units. In some cases where the day load is exceedingly light, it will not warrant the operation of the plant at all the load costs more than it will bring in. In such cases there can be no question as to the economy of the storage battery, since it can be charged during the night by power that would otherwise be practically wasted, and during the daylight hours it may carry the whole load. In such cases it will effect an economy in another way also, viz., by carrying up to its capacity the peak of the load. The economy in this direction will be the greater the sharper the peak, and will disappear as this flattens out. But the manner in which the storage battery is most frequently employed to advantage in a central station is by changing the load line from the irregular one, due to the natural load, to one which would be formed by a series of rectangles which would in the aggregate have about the same area. I have endeavored to illustrate this in the accompanying diagram. The irregular black line represents the natural load line. The spaces between the horizontal lines represent the units into which the equipment is divided, and the vertical lines the time at intervals of two hours. Beginning at 6 A. M. the load is a little above the best output for three units, but without too much. overloading these three units can carry it. The line gradually rises, however, until at 8 A. M. it becomes too much of an overload for three units. If a fourth unit were thrown in it would at first have to operate at only about one-third load, and, hence, very uneconomically. A storage battery load sufficient to make up a full load for this unit would, therefore, be thrown in at 8 A. M., and the new unit would thus work economically from the start. The part of the load furnished by the battery is represented by the shaded portion outside the natural load line—that is to say, this much energy is being absorbed by the battery and will be available at another time-less the loss due to inefficiency of the battery. I have assumed a battery efficiency of seventy-five per cent. We see from the diagram that the battery is charging 8 A. M. to about 4 P. M. Twenty-five per cent of the energy thus stored will, therefore, be lost. It will be apparent that this bears no relation whatever to the total amount of energy being generated at the time, for units one, two and three are working directly into the feeders, and are not affected at all. The energy thus lost bears an exceedingly small ratio to the total energy generated. At 4 P. M. the natural load consumes all the energy of the third unit, and in about twenty minutes has increased so as to be an overload. Now it depends upon circumstances whether we will bring our stored energy to the assistance of the fourth unit to carry the increasing load further or not. For the purpose of illustration, I have assumed that we will not, but will throw in our fifth unit, and complete its load by charging the battery. At ten minutes to five the load line crosses the fifth unit line, rising very rapidly, and the same question arises again, but I throw in my sixth unit at once, completing its load; the shaded area outside the natural load line again being the energy taken by the battery, seventy-five per cent of which will be available. At 5.30 the load exceeds the normal capacity of all six units, but the diagram represents them as carrying the increasing load until 6 P. M., when the seventh unit is thrown in and the load completed by the storage battery until 6.40, when the natural load requires the whole attention of all seven units. Now, the question arises, will it be necessary to add first an eighth and then a ninth unit to take this peak? If so, it will add considerably to the fixed charges on our power account. Will it be cheaper to supply battery capacity to carry this peak? That is a question to be decided separately for each individual case. Storage batteries will stand an abnormal discharge for a short time without serious injury. Some will stand it better than others. If it be too rapid, the depreciation charge will be increased, also if this abnormal discharge be continued too long. Then, too, at rapid discharge the battery becomes less efficient and the loss will become greater than twenty-five per cent. It is only by balancing these losses against the advantages that the question can be decided whether it will be more advantageous to throw in an eighth unit and let the battery take care of only the tip of the peak beyond that, or let the battery take the whole of the peak beyond the seventh unit line. In the diagram, I have assumed that the latter is more advantageous. The battery, therefore, discharges until 8.20, when it begins to recharge, supplementing the seventh-unit load until 9 P. M., when the seventh unit is thrown out entirely and the battery carries the remaining seventh-unit load until 9.25, the battery load being represented by the unshaded triangle within the load line. And so the battery and units continue supplementing each other throughout the remaining hours. In this way the battery will prove economical with some curves; in others it may not, but whether or not it will depends upon so many contingencies that it is utterly impossible to state á priori. There are other considerations than economy that might balance its absence, such as convenience as a regulator of potential; the facility it affords for distribution to sub-stations at high potential, resulting in an economy of copper; and the subsequent transformation down for local distribution; and some others others which will suggest themselves. Some of these might be controlling in special cases. Then, on the other hand, the employment of the storage battery is limited entirely to the storage of |