Popular Science Monthly/Volume 21/September 1882/Electric and Gas Illumination







THE period of contest and denial over the question of the possibility of producing a light of low intensity by means of electricity, that would be suitable for the general purposes of interior lighting, has about drawn to a close. It is now pretty generally conceded—what there has never been any reason for denying—that the known laws of electric transmission interpose no bar to the successful solution of the problem, but that the difficulties in the way are solely of a practical kind. And it is, further, quite generally agreed that these practical difficulties have been for the most part resolved, and the question reduced down to one of cost simply; and, while a good deal of discussion has taken place upon this point, but little has been written that will enable the general public to form a judgment upon the subject, and arrive at a trustworthy opinion of the relative cost of it and gas under actual commercial conditions.

In estimating the relative cost of the two illuminants, it has been common to compare simply the cost of the materials consumed in their production, or, when the cost of the apparatus necessary to generate the electricity has been taken into account, this has usually been upon the basis of a limited production, and, to this extent, unfair to electricity. A comparison, to be of any value, should be between plants of a size sufficient to reduce the cost to the lowest point at which it can be commercially maintained, and should include all of the items entering into it. The attempt has been made, in the following pages, to institute such a comparison, and present the facts in the case as they are, so far as they can be obtained. The comparison is upon the basis of works capable of producing a million feet a day, as, in such works, gas can be made as cheaply as in any that are larger. The figures for the electric plant are based upon the work of Mr. Edison, as he is the only one who has so far made any attempt to put in an electric plant upon an industrial scale. And, for that reason further, only his system of distribution is considered, though it may be a question whether it is the one which will prove most satisfactory in practice. An objection to it of considerable force in the opinion of some, is, the difficulty of handling engines and boilers with sufficient rapidity to meet great and sudden variations of demand, such as not unfrequently occur during the seasons of the year in which the weather is changeable. The variation that experience has shown takes place at different periods of the day can be met readily enough. On this account, and on account of the greater freedom secured in the matter of working various pieces of apparatus without interference, it would seem that the system of distribution which includes a storage-battery would be preferable, and may, perhaps, become the final form, adopted in electric installations. It can not well enter into the present calculation, however, as there are no data with reference to the first cost and depreciation available, and because the present secondary batteries do not seem to have yet reached a satisfactory commercial form.

The cost of such a plant for coal-gas will vary in this country from $2,500 to $4,000 for each million feet of the yearly make, but $3,000 may be taken as a fair average. Owing to the great variability in the demand for light at different seasons of the year, a gas-works of this size will be called upon to furnish but 200,000,000 instead of 365,000,000. feet a year. The plant will therefore cost $600,000. Of this, $250,000 may be taken as the cost of the mains, which, in average conditions, will have, for a works of this size, a total length of fifty miles, covering a district of about three square miles. To compare an electric with a gas plant, it is necessary to know the number of five-foot burners that will be maintained at the time of greatest consumption, as on this depend both the amount of horse-power required and the size of the mains to transmit the current.

The variation in the demand for light from hour to hour, as it would occur in average conditions on a bright December day, is exhibited in the following table, in percentages of the total make for the twenty-four hours:

7-8 a. m. per cent. 7-8 p. m. 12 per cent.
8-9 " ½ " 8-9 " 12 "
9-10 " ½ " 9-10 " 10 "
10-11 " ½ " 10-11 " 6 "
11-12 " ½ " 11-12 " 5 "
12-1 p. m. ½ " 12-1 a. m. 17/10 "
1-2 " ½ " 1-2 " 17/10 "
2-3 " ½ " 2-3 " 17/10 "
3-4 " ½ " 3-4 " 17/10 "
4-5 " " 4-5 " 17/10 "
5-6 " 16 to 20 " 5-6 " 16/10 "
6-7 " 14 " 6-7 " 17/10 "

This shows the time of greatest consumption to be between the hours of five and six, and the demand as high as twenty per cent of the entire daily make. In the case of the plant under consideration, the maximum number of burners that will have to be maintained at any one time is therefore 40,000.

Before proceeding to estimate the cost of the plant to generate and distribute electricity sufficient to maintain this number of burners, a few words descriptive of Mr. Edison's system will be desirable, especially as there appears to be considerable misapprehension on the subject. The distribution is what is known as in multiple arc—that is, the lamps are placed upon cross-wires between the conductors. Imagine a ladder erected upon an ordinary railway, so that it stands across the track, each foot resting upon one of the rails. Then these rails will represent the outgoing and returning street-conductors; the sidebars of the ladder, the house-conductors; and each rung, a lamp. The dynamo-machines generating the current are arranged in exactly the same way with regard to the circuit, all the positive poles being joined to one main conductor, and all the negative ones to the other. The arrangement is what is known, in the case of electric batteries, as coupling for quantity, as opposed to coupling for intensity, and is similar, to that of a number of pumps discharging water into a common main. This disposition of the electric-producing apparatus has the important advantage that the reserve plant, to meet contingencies, needs to be but a fraction of the total one; while, if each machine supplied an independent circuit, the plant would have to be in duplicate. As is well known, the steam-engines driving the dynamos are coupled directly to the machines, without the intervention of belts or gearing, the combination being termed the steam-dynamo.

The street-mains consist of wrought-iron tubes about two inches in diameter, containing two half-round copper rods imbedded in an insulating resinous cement. A main of this kind is carried continuously around each city block. At the intersections of the streets the conductors are brought together and joined to a main somewhat larger, termed a feeder, which supplies the current to these four blocks. It will thus be seen that the system of mains and the mode of production of the electricity are as readily capable of expansion to meet increased business as in the case of gas. The mains can be tapped anywhere for new consumers, and to meet this increased demand it is only necessary to run a feeder to the place of enlarged consumption, and increase the producing plant sufficiently.

What, then, will be the cost of such an electric plant to do the same amount of lighting as the above gas plant? If we take eight sixteen-candle lamps, maintained throughout the whole system for each actual horse-power applied to the dynamo-machine, engines with a normal capacity of five thousand horse-power will be required to sustain the maximum number of burners. This will include the reserve plant, as engines of a normal capacity of forty-two hundred can readily be forced to five thousand horse, or twenty per cent, to meet this extreme demand, and, with the generators arranged after Mr. Edison's plan, this per cent is an ample reserve. The maximum demand can, of course, be met either by forcing, or by running the entire plant at its normal rate, and forcing only in case of accident. To cover a district of three square miles, two distributing stations will be sufficient. The steam-dynamos may be taken as of two hundred horse each, working normally. The present steam-dynamos are of but one hundred and twenty-five horse, but they can be made two hundred horse with but slight increase of cost, which Mr. Edison contemplates doing in future installations. This will give thirteen steam-dynamos to one station and twelve to the other. These may each be placed at $8,000, making a total for the two stations of $200,000. That this is. a sufficient allowance will be evident upon considering the machines in detail. There are first the two hundred horse-power engines. No one will question that these can be obtained by a large buyer at $18 per horse-power, or $3,600 each.[1] This leaves $4,400 to cover the cost of the dynamo. The material in these, as now being constructed, is as follows:

Iron (wrought and cast) 40,700 pounds at 3 12 cents $1,425 00
Zinc (cast) 680 " 6 " 40 80
Copper 3,440 " 28 " 963 20
——— —————
44,820 $2,429 00

This leaves $2,071 for the cost of construction, which will be recognized as more than enough, when it is remembered that the cost of the iron as above given includes its shaping, and that the copper on the armature is in the form of bars and disks, which, with suitable tools, can be expeditiously constructed.

Adding twenty-five per cent to the cost of material for the 200 horse machine, there is still left $1,364 to be expended in construction. It seems to me, therefore, that $8,000 is a safe estimate of the cost of such steam-dynamos. Regarding the boilers, the sectional or water tube boiler, on account of its freedom from dangerous explosions, the smaller space occupied by it, its higher efficiency, and less cost for repairs, is in every way the best suited for a purpose of this kind. Such a boiler set ready for use, including stack and apparatus for handling coal and firing, will cost $20 per horse-power. The total boilers would therefore cost $100,000, making the entire producing portion of the plant, exclusive of real estate, $300,000.

As the Edison mains are now being laid, they will transmit a current sufficient to maintain from sixteen thousand to eighteen thousand sixteen-candle lamps. Taking the former figure, this is one and a quarter mile per 1,000 lamps. Basing the calculation for mains upon this mileage and the size of the present mains, the same number of miles of electric mains would be required as for gas. The present conductors are, as stated, in the form of half-round copper rods, of varying sizes, diminishing of course as they proceed from the station. They are, however, equivalent to round rods with a uniform diameter of one half inch. Such rods weigh 7551000 of a pound per foot, and 3986·4 pounds per mile, costing, at 28 cents per pound, 81,116 per mile. As there are two rods in each main, the cost per mile for copper would be $2,232. To this must be added $1,200 per mile for wrought-iron tube, boxes at the joints between the mains and house wires, and insulating material, and $1,000 per mile for laying, making the total cost of the main per mile, laid ready for use, $4,432. Four fifths of the mains would be of this size, the other fifth being feeders equivalent to round rods three fourths of an inch in diameter. These latter weigh 1·69 pound per foot, and would therefore cost $2,340 per mile, and, taking the cost of inclosing tube, insulation, and laying the same as above, their total cost per mile would be $7,196. The total cost of the mains, forty miles at $4,432 per mile, and ten miles at $7,196 per mile, would therefore amount to the same as the gas mains, viz., $250,000. If real estate be added at $50,000, which in most cities requiring this size of plant would be ample, the total cost of the electric plant would be the same as one for gas.[2]

The elements entering into the cost of the light to the company furnishing it are, in each case, the interest on the investment, depreciation, or the amount spent each year in keeping the property in good condition, the labor of all kinds—in the manufacture, distribution, and management—and lastly the cost of the materials used in its production. In the case of gas but a few of these items as they occur in American works are obtainable, so that recourse must be had to the published reports of foreign companies, and the like items estimated for this country. Of these, the reports of the London companies as analyzed by Mr. Field will best serve for the purpose of the present comparison.[3] Taking first the item of depreciation, we find that for the four metropolitan companies this was, for the year 1880, on the producing portion of the plant 9·86 cents per 1,000 feet of gas sold, or about five and a half per cent on the cost of this part of the plant as it has been taken in this paper. Calling this ten cents a thousand feet, we have $20,000 a year as the expenditure under this head, which is probably well within the actual figures of most American works. In the case of the electric plant four per cent is a sufficient allowance for the same item, which gives a yearly charge of $12,000, and a cost of six cents per 1,000 feet.

Depreciation of this part of the plant varies but little with different works, as the conditions upon which it depends are relatively constant, but that of the mains is, on the other hand, exceedingly variable. In a dry, open soil, gas-mains will last a great length of time, and even when they become entirely rusted through they will still continue efficient if undisturbed. They do not, however, remain undisturbed, so that in the most favorable conditions some expenditure is necessary to keep them in working condition. We shall probably not be far wrong if we take this at two per cent of the entire cost of the mains, which includes, of course, that of laying them. This item then becomes in the case of our gas plant $5,000 per year, and 212 cents per 1,000 feet. In the case of the electric mains, this percentage must be reckoned only upon their cost, exclusive of the copper, as this latter is practically indestructible, and can be used again and again. The amount upon which to reckon the two per cent depreciation is therefore $2,200 X 50$110,000, and the yearly charge $2,200, which gives 1·1 cent per 1,000 feet. The interest on the investment is the same in each case, and amounts to $24,000 a year, at four per cent, and to 12 cents per 1,000 feet. These items include all that are properly chargeable to the expense account of the plant save taxes, which would be about the same in each case, and which maybe neglected for the present. The plant account, then, stands, in the two cases, for each thousand feet or its equivalent:

Gas. Electricity.
Interest 12 · 12 ·
Depreciation of producing works 10 · 6 ·
"of mains 2 ·5 1 ·1
——— ———
Total 24 ·5 19 ·1
19 ·1
Balance in favor of electricity 5 ·4

The items entering into the cost of coal-gas are, exclusive of management, rent and taxes, etc., the cost of coal, of manufacturing, and of distribution. Taking the last first, we find 4·4 cents per 1,000 feet as the cost of this item for the four metropolitan companies. Putting this at 5 cents for American works, and deducting from this 212 cents for the depreciation of mains, which is included in this charge, there is left 212 cents for the cost of the labor of inspection of meters, etc., which constitutes the charge of distribution, and which would be about the same in both systems.

As the depreciation of the mains is not given separately, this item is liable to error, due to a wrong estimate of such depreciation, but, as it affects both systems similarly, it will not vitiate the results. Under manufacturing, the English report includes purifying, salaries, the wages for carbonizing, and wear and tear, which latter item has already been carried to the plant account. The first of these amounts to 1·82 cent; the second ·82 of a cent, and the third to 7·16 cents, making a total of 9·8 cents per 1,000 feet. This is probably much below the actual amount paid for these items in American works, but I am assured on excellent authority that, in works constructed after the best modern models, purification should cost the gas company nothing, and that all labor in the manufacturing department should be covered by an outlay equivalent to one man's wages ($2.50 per day) for each 40,000 feet of gas made per day. As the same amount of labor would have to be paid for each day in the year as on the days of greatest demand, this would amount, for a daily make of 1,000,000 feet, to 25 men whose wages at $65 per month (26 X 212) would be $19,500[4] a year, or 934 cents per 1,000 feet of the actual make. Including the cost of purification, and calling the amount 12 cents, we shall not be far wrong, or at least shall not exceed the actual outlay in the average works of this size. In the case of electricity the labor required at each station would be:

One chief-engineer $125 per month.
Three assistants (at $75) 225 "
Five firemen (at $60) 300 "
Total $650

—making $15,600 a year for the whole manufacturing plant, and 7·8 cents per 1,000 feet. To this may be added 115 cent to cover salary of electrician and incidental labor, bringing the item up to 9 cents.

There remains to be considered the cost of the coal in the case of gas, and the expense of running the engines in the case of electricity. The cost of coal per 1,000 feet of gas made was, in the case of the London companies, 3686100 cents, corresponding to $3.51 per ton, the make of gas being for this amount of coal 9,529 feet. This was offset by the sale of residuals, as below:

Coke and breeze 11·16 cents.
Tar and products 7·13 "
Ammonia and products 5·72 "
Total 24·06 "

—which leaves 12·8 cents as the net cost of the coal.

Compared with foreign companies, both in England and on the Continent, but very little is done with the residual products in this country, and the amounts received vary greatly between different works. Reliable data on this point can not be obtained, but under the most favorable conditions this item can not be taken as amounting to more than one half the cost of the coal, while with most works it is probably inconsiderable. The average price of the coal used may be placed at $4.50 a ton, and the amount of gas produced 10,000 feet, making the cost 45 cents per 1,000 feet. This make of gas can hardly be maintained with a production of residuals equal to one half the cost of the coal, but. assuming that it is, the cost of the coal becomes 22 cents per 1,000 feet.

In the foregoing estimate of the electric plant, it has been assumed that eight lamps could be maintained throughout the entire distributive system for each actual horse-power expended upon the pulley of the dynamo-machine. That this is entirely feasible has been proved by careful tests made by experts in no way interested in any of the lamps, and their results can therefore be accepted without question. For such a use as electric lighting, the cost of a horse-power may safely be taken as not above the best results hitherto obtained in practice. In general manufacturing, the item of power, while important, is not sufficiently so to demand that constant and great care necessary to obtain the very best results, and hence few engines and boilers yield in practice the same results as in special tests. With electric-light companies, this item, on the contrary, is vital, and we may confidently expect to see them in time obtaining their power at a considerably less cost than is now common. Mr. Edison finds as a matter of fact confirmed by several months' test at Menlo Park, that he is able to maintain a horse-power an hour with five pounds of slack (one third pea and two thirds dust), costing $2.45 a ton. For the purpose of the present comparison, however, it is best to make a liberal allowance, and take for a 200-horse-power engine a consumption of four pounds of coal an hour, the coal costing $4.50 per ton of 2,240 pounds, delivered. A horsepower will then cost 810 of a cent an hour, and we may rightly abate our liberality sufficiently to include in this the cost of the oil for lubricating the engine and dynamo.

The maintenance for an hour of 200 electric burners, the equivalent of the 1,000 feet of gas, will therefore cost 20 cents, as against 2212 cents for the gas.

Summing up the results so far obtained, the two accounts stand as follows:

Plant Account. Per 1,000 Feet.
Gas. Electricity.
Interest 12 · 12 ·
Depreciation of producing works 10 · 6 ·
Depreciation of mains 2 ·5 1 ·1
——— 24 ·5 ——— 19 ·1
Manufacturing expenses:
Labor 12 · 9 ·
Coal 22 ·5 20 ·
——— 34 ·5 ——— 29 ·
Working expenses:
Distribution 2 ·5 2 ·5
——— ———
Total 61 ·5 50 ·6

Under this last heading there should be added rent and taxes, management, law charges, bad debts, and various incidentals. These can not be separately arrived at with any closeness, but they may be taken in the lump as about the same part of the total charges as in the case of the London companies, which is 16 per cent, exclusive of the interest on investment. This in the present case would be 9*4 cents per 1,000, bringing the total cost per 1,000 up to 71 cents with gas and 60 cents with electricity.

The promoters of the electric light would probably demur to this statement, so far as rent and taxes are concerned, as they insist upon the much smaller real estate required with the electric than with a gas plant. This difference does not, however, seem to me sufficient to be of any practical moment, as the real estate in the case of electricity is in the district supplied, where the price of land is relatively high, while the gas companies can readily place their works in such locality as to compensate in lowered land value for the greater amount required. Gas companies can, moreover, build within much smaller limits than usual when for any reason it is desirable, and closely approach the space requisite for an electric installation.

An item of considerable amount which has been omitted from the estimate for electricity is the cost of the renewal of the lamps. With the general introduction of incandescent electric lighting, this is a charge which would fall directly upon the consumer, but it is one which would steadily diminish with improvement in lamps. Assuming, however, that it is a legitimate charge upon the company supplying the light, the item amounts to 10 cents per 1,000, if the lamps have a life of 600[5] hours and cost 30 cents. This brings the electric account up to 70 cents per 1,000.

So far as coal-gas is concerned, then, these figures show a slight advantage in favor of electricity, and while they are only approximative they are near enough to the truth, I think, to represent the actual relation of the two illuminants. While very much doubtless remains to be done in the improvement of coal-gas manufacture, it does not seem probable that this will affect its cost of production to the same extent as future improvements of electric apparatus may be expected to decrease that of the electric light. Looking closely at the two accounts, it does not seem probable that the item relative to plant will be materially lessened in the future. The cost of the plant has already been taken at a figure very near the lower limit, so near that the substitution of this in its place would make a difference in the yearly plant account of but 212 cents per 1,000. We may, on the other hand, expect improvements to largely reduce the cost of the electric plant. On Mr. Edison's system of distribution, the size of the conductors varies inversely as the resistance of the lamps, so that they may be materially reduced if the resistance of these latter can be increased; while any improvements affecting the number of lamps per horsepower diminishes both the interest account by reducing the plant and the actual cost of production.

How far coal-gas can go in a reduction of the cost of production it is difficult to say, but I think the lower limit may safely be taken at the point at which the sale of residuals pays for the coal. Both of these items—cost of coal and prices of residuals—are practically beyond the control of a gas company. The coal is already purchased in the open market at the lowest figures at which it can be obtained, and the market for residuals depends chiefly upon the development of chemical industries, which can hardly be hastened by the action of a gas company. This market is a steadily growing one, and it is not impossible that the residuals will in time pay for the coal, though it is hardly probable. The items of labor and distribution can not probably undergo any considerable reduction. The limit, then, below which it does not appear that there is any probability of coal-gas falling in this country is 46 cents per 1,000, which is a figure that may be reached by electricity without assuming anything less, probable than the above supposition respecting gas. It is only necessary to get ten lamps per horse-power, and produce the latter with three pounds of coal an hour, to bring the cost down to 47 cents, exclusive of the lamps.

As a present competitor, however, what is known as water-gas—gas produced by the decomposition of steam in the presence of coal or oil—appears to be the more formidable. This mode of gas-manufacture has the advantage of coal-gas in a lessened cost of the producing plant, a smaller labor account, and a decreased depreciation of the generating apparatus. Its successful competition with coal-gas ultimately depends upon what the latter can make of its residuals, as there is no offset of this kind in its case, but with present conditions it can go below it. The producing portion of the plant costs but little more than half that for coal-gas, while the labor is about a third, and depreciation but slightly more than this. A sixteen-candle gas will require three gallons of oil per 1,000 feet, and can be made with oil at 5 cents a gallon and coal at 84.50 a ton, at an expenditure of 28 cents per 1,000 feet for materials. The total cost will not exceed 60 cents.

Such, then, appears to be the relation of these two agents on the basis of illumination solely, but it must not be forgotten that the amount of light which each plant can furnish does not represent the actual relative capacity of the two. The electric plant can be run not only four hours a day for light, but any further number of hours for power, without any increase of the machines. The gas-plant, on the other hand, would have to be increased, to furnish both power and light. That this advantage of electricity is liable to be a very important one will hardly be questioned, when the extent of the field open to electro-motors is borne in mind.

On these figures the cost of electricity is near enough to that of gas to enable it to offer a very substantial competition, and one which may be expected to grow stronger with increased experience and future improvements. That under the stimulus of this competition considerable improvement will be made in lighting by gas seems very probable. Already it has been shown that in the matter of burners there is a wide field for invention, and that the results now usually obtained are much under what are possible. With the high-power burners of Siemens, the illumination obtained from sixteen-candle gas has been more than doubled, and in others it has been carried up to from five to five and a half candles per foot. How suitable burners yielding such a great increase of light will be for the general purposes of lighting, and whether they can with advantage displace the simple flat tip, remains to be seen, but the present indications are that it is chiefly through the use of improved burners that gas must endeavor to resist the assaults of the incandescent light. Competition on the basis of a gas of higher illuminating power simply, without a resort to improved burners, does not seem very promising. The recently published report of the sub-commission, appointed to test the incandescent lamps at the Paris Exhibition, of which Mr. Crookes was a member, shows that a thirty-two candle lamp can be maintained with an increase of from 28 to 37 per cent of the power required to sustain one of sixteen candles, while with gas such an increase of illumination will require an additional expense of fully 50 per cent of the cost of one of the lower candle-power. This is so with the Lowe gas, with which three gallons of oil are sufficient to give sixteen candles, but six are required for thirty-two, and it is not probable that coal-gas can be enriched any cheaper. Whether the limit to progress in gas-lighting—both in the matter of improvement of manufacture and burners—is sufficiently far off to give gas unquestioned possession of the field of lighting or not, the result can alone determine. But, if the figures presented in this paper can be at all relied upon, they show that gas manufacturers and those interested in gas property will do well not to underrate the strength in their own domain of this rising industrial power.

  1. Mr. Edison informs me that engines of 200 indicated horse-power are being purchased by him for $1,750 each, delivered in New York. This estimate is, therefore, much too high, but, as the comparison of plant in the text is based upon it, I have thought it best to let it stand, and point out the needed correction here.
  2. While this estimate seems to me not far from the expenditure that would be actually required for this size of plant, it should be stated that it is lower than any of those given by the electrical experts examined by the select committee of the House of Commons in its consideration of the Electric Lighting Bill.
  3. Having been unable to obtain a copy of Mr. Field's "Annual," I have taken the figures as quoted from this for the year 1880 by Mr. Dowson, in a recent lecture before the Society of Arts.
  4. The engineer furnishing the information on which this statement is based informs me that this should be $12,500, or $2.50 per 40,000 feet of the actual yearly, instead of the maximum daily, make. This would reduce the item 934 cents to 614 cents per 1,000 feet.
  5. I am informed by Mr. Edison that the average life of the lamps is now 900 hours, including 3 per cent breakage in handling.