of the American Institute of Electrical Engineers, Eliot, Me., July 27th, 1897. President Crocker in the Chair.

THE COST OF STEAM POWER.

BY HORATIO A. FOSTER.

In this age of electricity, when so much is being done in the way of developing its applications, and especially in its application to the transmission and distribution of power, perhaps no one item causes so much controversy as the cost of steam and other familiar forms of power. It seems to have been taken for granted for years that anyone and everyone knows the cost of developing power by steam, and in almost all cases when a person stating such costs is pinned down to facts, it has been found that he is either guessing, or quoting those he supposes are authorities. It is to be presumed that statements of actual tests for cost might have been published, but the writer has in a somewhat extended search run across but one such statement that might be considered reliable, and that was a test for something under 100 days, and that practically confined to the warm season of the year and not therefore of the greatest value.

Other estimates have been made from extended experience and carefully revised theory, and in fact little can be added to tables published during the past few years by some of our most eminent and best qualified engineers, in the way of carefully prepared estimates of what should perhaps be called the average cost of producing steam power under various conditions as to amount of power and the type of machinery employed.

It must be said, though, that almost all these estimates have been made for powers of large size and where little variation of load takes place, and, while such estimates are of great value for use in discussion of the subjects, the writer has found them somewhat unsafe to use unless their results are applied in con

nection with extended experience, for, while the results considered as averages are probably correct for ordinary conditions, they cannot of course be expected to include extraordinary conditions, and the results of this investigation seem to show that individual cases vary so much that it is not fair, and may be misleading to state average costs.

It seems a pity that the cost of power is most often quoted for one horse-power for one year, and so freely assumed that there is an average cost of power on which one can base estimates in the cost of manufacture; for the writer will endeavor to show, as this paper progresses, as others have done, first, that any general statement of the cost of power per horse-power per annum is incorrect, for the reason that such a statement does not ordinarily take into account the number of hours run per diem, or the number of days per year, and even in that class of plants supposed to run, say 308 days per annum, it is seldom one can be found that runs a full year, and no two run the same length of time; and second, that no such statement of an average cost of power is of any value whatever, unless it be for plants employed in similar work, and under exactly similar conditions, and even then the amount of power developed is such a factor for varying the unit cost that it is in the opinion of the writer quite fair to say that there is no average unit cost for steam power, that is safe for general application, and it is thought that a somewhat careful study of the tables and computations accompanying this paper will help to a clearer understanding of the reasons therefor.

Although in the great majority of instances office accounts are so kept as to give no real basis upon which to compute the cost of power, and in most off-hand estimates, the rated capacity and not the average power developed has been used as a divisor, there can be no doubt that in some of the older industries, say in the great flour mills and the large eastern cotton factories, where prices of manufactured goods have been tending downwardly for so many years, and very close accounts of unit cost have necessarily been kept, and, as in all other departments the power has been very carefully looked after, that in such cases it is very likely that careful comparisons will show the cost of power to be more or less uniform when used under similar conditions, but reference to the tables will show this to be more widely at variance than is commonly thought.

It will be the endeavor of the writer to show by tabulated data as far as possible, the condition of the plants tested, without going any more into detail than seems necessary for a clear understanding.

In order that the value of the results may be determined, it is necessary to give some description of the methods pursued in obtaining them, and, as they were as far as possible the same throughout, a general description will do for all except the special cases.

In order that the greatest value might be obtained, it has always been the opinion of the writer that no results for a period short of twelve consecutive months should be used, and in all but one case which was for eleven months, a year has been taken as the smallest unit of time.

In a general way, the unit of time having been settled, where possible a unit of output was selected, after a careful and somewhat extended study of the special business in all its conditions, on which averages could be constructed and which could be turned into horse-power after determining its rate.

Indicator tests were always made both for determining the rates of power and the ratio of power to the unit of output: for instance, in a flour mill it is only necessary to apply indicators to the engine cylinder for a certain period of time, not less than 24 hours, and to count the actual output in barrels, to determine a fairly accurate ratio of horse-power per barrel of production, after which it is a simple matter to determine the average for a year from a total production and the actual time during which the mill was running and the output recorded; or in an are lighting station, if an indicator test is made of the engine for the entire time per day, during which lamps are burned, and the number of lamps running during the same period be recorded, then also it is easy to determine the ratio. of power to lamps, whence by computation from the records kept of the number of lamps burning and the time in hours during which they were lighted for a year, the lights can be reduced to horse-power, and if accurate and well classified records of expenditure are kept, the annual and hourly cost of power is determinable to a close degree of accuracy. Of course, wide variations in the amount of power used will often change the ratio, but care has been taken in these tests, not to base them on periods of changing power consumption, and it is therefore

thought that the results are within the ordinary limits of accuracy.

For the sake of uniformity, as well as to reduce the items of account to as small a number as was considered consistent with clear understanding, a list of cost accounts was determined on, that in most ways agrees with work previously done in this line, and which it was thought could be most easily met by the most ordinary bookkeeping.

The list is as follows: Operating Expenses.

Fuel.

Wages.

Supplies.

Repairs.

Water.

Incidentals, including lighting and removal of ashes. Fixed Charges.

Interest on cost of plant.

Depreciation on value of plant.

Insurance.

Taxes.

In considering the value to be given to each item, many things had to be discussed in order that proper separation of expenditure due to items other than for power could be made, as for instance the item of fuel; in many cases buildings have had to be heated, either by exhaust or live steam; steam has been used for hot tables, as in the case of newspaper offices, and for dry rooms for lumber etc. as is the case of a furniture factory. In all such cases the study of the plant, consultation with the engineer in charge and the management, coupled with the experience of the writer has ensured the determination of an amount satisfactory to all concerned as being the nearest correct obtainable under the ordinary conditions of practice.

In the item coming under the head of fixed charges, a somewhat different treatment obtained, as so little attention is given these items by the ordinary business management that in most instances not only the value of the plant, but the items going to make up the fixed charges had to be determined by the writer from his experience, together with such information as could be gotten from the people in charge. While insurance and taxes were of course determinable from the books, interest was rated in almost all cases at 5 per cent., and depreciation at rates varying

from 2 to 10 per cent. for building, and 5 to 15 per cent. on steam plant, all according to type of material in use, its condition, the manner in which it was handled, etc. In but few cases were either of these items considered in the estimates of cost by the owners. It must be remembered in this connection, that fixed charges go on all the time, 24 hours per day for 365 days, and that the unit cost is the same if computed by the year, but does not have the same relative value when computed by the number of hours the power was actually in use. In passing, it may be well to say that the number of firms in good business standing that are entirely at sea regarding the actual cost or present value of power plant is to say the least astonishing.

One item taken into consideration in all the tests, and one seldom considered at all in the ordinary statements of the cost of power, is the value of the building or part of the building devoted to the use of the power plant. In some few cases buildings were set apart for this purpose, and where the value was known it was of course easy to fill out the item; but in cases where the power plant was in the same building with the other portions of the plant, it became necessary to set off the value from the rest, and this was most often done by determining the renting value of the room used, which can usually be done with small chance of error, and then to capitalize the rent as the value of that part of the building devoted to power purposes.

In one item used in these tests, the writer apparently differs from most other estimators of the cost of power, and that is in the use of the indicated horse-power rather than brake horsepower for the divisor, and his reasons for using it are as follows: viz:-Indicated horse-power can nearly always be determined to a degree of accuracy only limited by the usual errors of instruments and their users, generally conceded to be small; while brake tests can seldom be made in practice, and the net power is therefore most often determined by assuming a certain percentage of engine and shafting efficiency which under general conditions is quite incapable of proof, and the errors of which must be added to those of the indicator test. The writer therefore prefers to eliminate this chance for error, and should any one desire to change the figures here displayed to brake horse-power, it is only necessary to make the single computation due to any engine efficiency that he may wish to assume.

Referring now to plants tested, those on which reports were