Amazing articles on just about every subject...

The Production of Electricity by Steam Power

( Originally Published Early 1900's )

I CAME HERE tonight to keep a promise which was broken some six months ago. I was then called to Washington on some business of alleged national importance, which did not prove to be of great moment. These things are the troubles which beset us all in war times. To see so many people present at a war-time convention is a surprise and a great delight.

You are more interested in power delivered and available for your work than you are in the question of generating or producing power at the terminals of a dynamo-electric machine; therefore, it will be necessary for me to talk discursively, to get into questions of distribution and transmission, and to some extent into the utilization of power in order to get before you the ideas which I think you expect me to present, namely, the present status of the production of electric power from steam, as I understand that status.

I have not prepared a paper for you; I tried to avoid that sort of thing. It is one thing to talk to an audience on a subject fairly fresh and recent in your mind, and another thing to write a composition which will stand critical analysis in the office or the re-search laboratory afterwards. And I have an inherited national prejudice against reading an address anyhow.

You know the tale of the Scotch preacher whose early effort was a great failure, and he did not know why until he asked an old friend, who said, replying to the preacher's question of why he failed: "Robbie, there were three reasons: one was, you read it; the second was, you read it badly; and the third was, it was not worth reading."

I intend to avoid the first two reasons, and as to the third trouble, you may judge for yourselves later.

Power Generated and Delivered

Let me make the point again, that delivered power and generated power are two different factors. Generated power is the actual power produced at the terminals of your machine; the output is the amount of power delivered from your station into the trunk lines. The difference between the two represents uses in the power house, lighting, heating, excitation, the operation of auxiliaries, and all the minor losses incurred in any power plant operation. The net power output is what the great majority of central station men are concerned with. Then they have a third expression "cost of power delivered," and that is the amount of power delivered to consumer and presumably paid for. The con-fusion between these three is sometimes, I am sorry to say, intentional. Some persons have the habit of expressing their costs in generated units. In one instance I found a discrepancy between generated units and units delivered to the lines for distribution use, of the order of 10 per cent. And when that error was corrected the costs boasted of were not worth looking at twice. Costs said to be remarkable proved to be only good average practice.

The final figure, delivered power, varies again a great deal. In a close network, such as serving one big mill owning its own power plant, the difference between output and the actual delivery to the points of use may be 8 or 10 per cent. In a complex system, covering a large area, and delivering power in different forms, the order of difference is more likely to be 20 per cent and upwards. That is to say, if the customers' meters of a large central station network account for 80 per cent of the output from the power house, they are doing very well. That distribution efficiency is first-class practice.

There are many well operated networks, which, from obviously unavoidable conditions, do not have a distribution efficiency exceeding 72 or 73 per cent. Commercially their practice is all right, but the difference between power generated and power delivered by the central station systems to the customers' terminals is large enough to be an appreciable factor in the cost.

A Bit of History

To come here and talk to this audience one should qualify as a teacher. Once in my lifetime I qualified as an expert witness, and I hope the Lord will forgive me for it. It was somewhere about 1870- I am quite as old as that—in 1870 I made my first acquaintance with the inwardness of electric power. My text-book was one which I doubt if any one here would accept. It was Dr. Lardner's Museum of Science and Art. I read that very earnestly as a child, and therein I made the discovery that James Watt did not invent the steam engine. And I made a further discovery which many of my good friends have not yet discovered; which is that what he did invent is the separation of the condenser from the cylinder. Some-how it seems to me that every development of efficiency in steam power from James Watt's day to our own, has been in further separation.

The first division in the steam chamber was made by James Watt; the separation of the steam chamber into successive chambers, all the way from double to sextuple chambers, is within the memory of most of you here; and the establishment, instead of successive chambers, of a long gradient, as in the turbine, is a matter of the last fifteen years. We are now separating the functions of the condenser into condensation of steam and removal of air.

Therein—in separation—has lain every vital improvement in the steam engine. The others have been, not in physics, but in mechanism.

The next date in my memory is 1888, when I got acquainted with a certain 12 by 12 high speed engine running 280 or 300 revolutions per minute, in Southern Illinois, in which, and in the electric system connected with it, my then employers had a 50 per cent interest. And the problem then put to me, I being an alleged electrical engineer—I say "alleged" advisedly—was why if an engine of that type would produce on test a horse power for every 4 pounds of coal, why that particular beast should re-quire 10 or 12 pounds. That involved a determination of what the engine was really doing, and it startled me to find that its consumption was in fact 17 pounds of coal per horse power-hour. There are a number of varieties of that engine, and it has many brothers today. You must blame the combination of boiler and engine, and then possibly throw in some of the human element. But when I got through I had a clear idea of what load factor meant, and what cylinder condensation meant. I became familiar with, and contemptuous of, automatic cut-offs, and to think that throttling had some virtues. It had, and still has.

The Willans Law—the law that the steam used by any engine is a constant plus a variable, the variable being proportioned to the load—was announced about that period, and was one of the first recognitions of what actually went on in a high speed engine.

Willans (you may remember) divided his very successful engines into many chambers. He reduced the wetted surfaces; he produced a wonderful valve, and those engines of his are still continuing in service today, and they are very hard to beat in steam efficiency, especially if given a little superheat.

I find that vision of load factor had become a rule, a guiding law, to me, when in 1894 I found myself called upon to design a plant for a street lighting load, a matter of 2,000 horse power, in four or five engine units, with, at night, a street light load of 4,000 hours a year, shut down at the coming of daylight; and only a trifling load all day.

I deliberately selected a marine type steam engine for 160 to 170 pounds steam, at practically a fixed cut-off, although there was a governor of a sort; and I met the occasion of a day load, which ran from 150 horse power down to 75 horse power, by the expedient of letting steam pressure run from 160 or 170 pounds down to 80 pounds on the boilers—down as low as it would go before I did any new firing. The result was it reduced the use of coal to 4 pounds per kilowatt-hour, which would be equivalent to 2 1/2 per indicated horse power.

In that design there was a point I would like to put before you, as it illustrates the cost of distribution. I and my associates deliberately chose series wound dynamos of rather low efficiencies, down in the eighties, as against bigger dynamos of better types, because the problem had to be treated as a whole as one of cost of current actually delivered to the street lamps; and since the circuits ran into some fifty square miles of territory to light, it was highly desirable to do as we did and keep transmission losses down to 7 or 8 per cent, instead of accepting higher power house efficiencies and let the other losses run to 18 or 20 per cent, which the higher power house efficiencies would have predicated. Please notice the acceptance of the lower efficiency at the power house, to obtain maximum delivery at the point of use.

In 1908, I believe it was, I made my first personal, real contribution to the reduction of steam power cost. Up to that time I was merely developing the ideas of other men. Then I came to the conclusion that a large boiler would be perfectly safe and feasible; that a very large furnace was desirable for burning bituminous coal; that stokers should have certain characteristics; and I was able to get the financial support to build a power house designed for 56,000 kilo-watts, in which the boilers and furnaces were of that type.

In 1912 two of these boilers were tested very liberally by Dr. Jacobus, and the results of those tests are in the text-books now. They established the very great importance of the correct adjustment of the furnace, and the stoker as a part of the furnace, to the proposed fuel and the proposed work. They also decided that the efficiency curve, of the boiler and furnace together, is a very flat curve; that a range (in this instance) from 40 per cent of the nominal rating to 200 per cent of the nominal rating of the boiler, was entirely practicable. That upper limit has now been much exceeded. A design for long range is essentially a compromise design. Please do not overlook that.

Connors Creek Power Plant

Lastly, in 1914 and 1915, came the Connors Creek power house, which represented the best I knew, and which is the occasion why I speak to you tonight. The figures of the operating cost of that power house have been given to the public. They have been published at six months' periods, and by years, and as far as I am advised they have not been, for long periods and for a complete power house, thus far improved upon. I know of three power houses in operation and some others coming through, which should beat them 8 or 9 per cent; but their figures have not been given out, and we are from time to time assured that the beating to which we are entitled has not yet been given to us.

The three Tables, I, II and III, are tabulations of the Connors Creek plant costs for three periods of twelve months with a six months' shift between periods. They thus cover two years. It is necessary to use figures for one year in such presentations because of the difference between summer and winter costs and because of occasional temporary aberrations of service, or of fuel, which would give exceptionally good or exceptionally poor showings for shorter periods. It will be noticed that the increased cost is almost entirely in the increased price of fuel. To a small extent it is due to uneven quality of fuel, because for the last nine months we had to burn any kind of coal we could get, and a change from one fuel to another is always temporarily accompanied by poorer combustion. The other causes affecting costs are very decided increases in hourly rates of pay, the effect of which is cancelled to a good extent by increased output; and the fact that, during the Table III period, No. 2 turbine was running without its last row of nozzles and buckets, and its steam consumption was thereby increased some 4 per cent.

Note that the stated output is kilowatt-hours metered to out-going 23,000-volt trunk lines, also that the stated load factor is a customary central station load factor based upon the maximum half-hourly reading of output during the period.

Let me say this: Fuel cost is the one figure on which comparisons can be made, and fuel cost must be expressed in heat units to make the comparison. These Connors Creek figures show year in and year out an approximation of 20,000 heat units per kilowatt-hour with periods below and periods above it. In cases where they exceed the 20,000 figure, the excess is of the order of 1 1/2 or 2 per cent, and when they get below it, it is at very favorable seasons of the year. They are the best which can be done with the present limitations of that power house.

The house is of a design, and a selection of machinery, of the years 1913 and 1914. It is not completed. When completed it will do a little better. The reduction of losses by radiation from the boilers proper is so extreme that the boiler room must be heated in winter for the firemen. The turbines are essentially 12-pound per kilo-watt-hour machines, and are habitually operated at their best points, say 11 1/2 on test conditions; but we do not rate them at that.

Heat Balance

The particularly interesting feature is the arrangement made for getting the most out of the steam used by auxiliaries; and here begins what I might call Chapter 1 of Instruction—because so far I have only given you historic classification—as follows:

Heat Balance: You shall not throw heat overboard beyond what you cannot save, and you shall get as much power out of all the steam you make as is possible.

Now, if you will turn your memory back you will recall that feed pumps and condenser pumps and auxiliaries of all kinds were once upon a time habitually driven from the main engine, and therefore they had the efficiency of the main engine. Marine practice, until recently, was exactly that way, and thereby rose a compromise, due to the limited spaces and weights in marine practice, of running at 24 to 25 inches vacuum; because experience showed that to put the feed water into boilers at temperature corresponding thereto was more efficient than to get better vacuum at the expense of colder feed water. Colder feed water meant you were pumping too much heat overboard; and that loss overboard is the one hopeless proposition today. You will remember that something like eight-tenths of the heat units are hopelessly lost in the condensing system; and therefore any heat process which uses electric heat originating in fuel carries a heavy initial handicap.

A mighty poor combustion of fuel—be it coal or oil or natural gas—should give 50 per cent efficiency of heat transfer into the work. I think the domestic gas range gives between 35 and 50 per cent efficiency when operated by a reasonably careful cook, who does not let the gas burn all night through neglect to shut it off when through using it; and an efficiency of 50 per cent of the delivery of the heat to the work from combustion, in a process of that kind, should be made every time, and higher efficiencies should be practicable. On the other hand, by our present methods you will never get higher efficiencies than that corresponding to the possible 100 per cent efficiency of the transformation of electricity into heat, multiplied by the possible 20 per cent efficiency of the transformation of fuel energy into electricity. Therefore, in electrical heating units, you use electricity because of electricity's other merits—not for its thermal efficiency from fuel to work. And its large use proves that wise men are willing to pay for those other merits, and that fuel economy is not the only thing to consider in industry.

Put me down as saying that where electric heating, electric steel making, electric reduction of ores are used and the origin of the electricity is in fuel, you are absolutely applying that heat wastefully, considered as an application of heat, and the wastefulness is hopeless. The loss overboard in the condensing water is the big loss. All other losses, such as stack losses and radiant losses from boilers and engines, are minor in comparison.

The losses in auxiliaries have been very great. You will recall that the next style of auxiliaries was in the use of separate direct-acting steam pumps, which were popular improvements in pumping devices; and we had the theory propounded that inasmuch as these were always operated with a closed heat cycle, the steam used in those auxiliaries was not a loss, but the heat therein was wholly returned to the system. That was true, but the heat was returned without removing anything like the amount of power which should have been removed.

The Connors Creek auxiliary plan (which as far as I am aware was not preceded by and certainly not taken from any earlier design, but was one of my own sketches, made in 1912) is to operate the auxiliaries by electric drive from a separate turbo-generator, exhausting into a condenser whose temperature is controlled ac-cording to the requirements of the feed water. That is to say, you run a separate electric generator of comparatively small size, and from that you drive your auxiliaries (except a steam turbine feed pump) and the exhaust from the turbine of that generator goes to heat your feed water to the economical temperature. That economical temperature being settled by other considerations, you get from the steam which goes to supply your auxiliaries, all of the possible power and likewise all of the heat; and your auxiliaries have the great convenience of the electric drive.

At Connors Creek we have a method which is capable of various and numerous changes. In our last sketch the feed pump is also made electrical, with a standby steam feed pump to meet occasional conditions. The combinations possible are legion. I could talk to a mechanically interested audience at least 3 hours on this subject. Right there let me say that all-steam drive of auxiliaries must be wrong; and all-electric drive usually is wrong. The correct choice lies between the two.

Let me say this about Connors Creek: There is no economizer there; that is a commercial proposition. The use of the economizer would unquestionably save 4 per cent of the heat units; and commercially, under the conditions of coal cost as they were two or three years ago, it would not pay. At the present time it would probably pay if the thing could be "wished into place," and the problems required to be solved in the all-steel economizer could be solved with dispatch. If coal stays at the present ordained figure—not to speak of the actual figures we are paying—the economizer will be warranted and we have saved space for it. Under the conditions of the past the economizer would not have justified its existence. If put in, the economizer would better the economy 4 per cent; then the turbines could be replaced by similar machines 4 to 5 per cent more efficient; and with the heat balance betterments we can see, we would pick up 1 or 1 1/2 more per cent. That is to say, we could reduce the present 20,000 heat units per kilowatt-hour to 18,000. That is an entirely practicable figure at present, without going into new design. There are certain ways whereby a still lower annual heat cost could be obtained in practice, and they are commercial, such as improved load factor.

Let me say also there is a difference between summer and winter efficiency, as in winter you have a much better vacuum. The colder water of our lakes and rivers in the winter months means that a better vacuum is practicable. On the other hand, if you over cool your condensate you are losing too much heat over-board. And the winter radiation losses from a power house, built as a power house ordinarily is, and not as a refrigerator plant might be, are a decided loss.


Now the total cost is the fixed cost plus operating. The division sometimes made is into standing and running costs—the standing costs being those which make the station ready for service, and the running costs being the added cost due to service or output. You get there an equation like Willans Law. Let me say this: The fixed element is so very large that at poor load factors it is controlling. With some of you the load factor is high; with others it is inclined to be low. The electrochemical load factor is generally recognized as specially high; it is of the order of 90 per cent of what would be obtained if the maximum hours, that is 8,760 hours, of maximum demand were obtained during the year. You gentlemen are in the habit of telling us you have 100 per cent load factor. Ninety per cent is going strong, and figures in the eighties are more frequent. On the other hand, a mixed service is doing well if it shows 40 per cent, which, expressed otherwise, would be that if that station could operate at full load for four-tenths of the time its output would be the same as it is, at the loads available, for the whole 8,760 hours of the year.

Any central station with the load factor above 40 per cent is doing well, indeed. Below that is common enough, The load factor of residence lighting is of the order of 8 per cent; which means that the costs for residential lighting service are nearly all fixed charges. You can almost forget running charges on residence business. Since it is the custom to charge according to kilowatt-hours used, the rate per kilowatt-hour for residence service is much higher. Did we charge according to the rate forms used for electrochemical service, the difference would be seen to turn on that one point of load factor. Residence service uses the equipment about eight-hundredths of the time that electrochemical service might use it.

Now, as to the capital cost, a fixed or first element: The capital cost of the Connors Creek plant, for the first units, is between $50 and $60 per kilowatt. If you buy land for 300,000 kilowatts you must prorate that over the proposed investment or installation; and you put in railroad tracks, and dike out lakes, and put in all the facilities and offices, when you build your first units of a big plant, and all of these cause the unit cost to go high; therefore the variation is between $50 and $60. But the crazy prices of the last two years would make the cost go between $90 and $110 per kilowatt. One hundred and eighty per cent is just about the cost of such construction now as compared to the cost we considered normal before the world went crazy.

Now, money for that sort of business formerly could be had at rates which did not very much vary.

Taxes on a public service plant are assessed on a liberal basis. We are very honest, we public service people. We always confess to a liberal and honest basis for taxation. Now that is no slight matter. Communities which in the past have favored the manufacturer with a low estimate on his plant, for taxation purposes, will assess a public service corporation up to the eyes; and let me say that taxes in the last year required us in Detroit to earn $1,600 a day, to pay taxes every day, Sundays included.

The cost of money, plus the depreciation reserve, plus the taxes means that the central station must figure on a return never less than 12 per cent, and sometimes exceeding 15 per cent. So your first figure, based on the central station method of obtaining your power supply, is to figure 12 to 15 per cent as a figure on an investment per kilo-watt, which was between $50 and $60 and which now is between $90 and $110.

Let me say a word here about depreciation: Now, the depreciation conditions are really affected by an electrochemical load. Depreciation, save and except "obsolescence," is higher with higher load factor. Depreciation will be some function of use, and a high load factor will obviously mean a higher rate of depreciation. That should be remembered in figuring costs of any purely electrochemical plant.

Effect of Power Plant Location

Now, assume we have the problem put to us by some of your members, that you want a power plant, the object being low total cost of power, and you have some choice of location. The first thing to consider is how the location may affect the power plant; and that leads to discussion of those nonengineering problems I spoke of earlier.

With electrochemical industries, of course, you must consider your raw material and your markets. You must have good ship-ping points; you must have fairly good labor conditions, etc. These conditions would have a large influence in determining the location; and if you leave us free to transmit the power over from a satisfactory location to the power plant—if that condition be in the premises of this discussion—we can forget these considerations for the present and consider one location against another as reflected in design.

242 Take item No. 1—the fuel cost. If the load factor is of the order of 90 per cent, fuel cost will be the most important thing to consider. In connection with fuel cost you must consider the certainty of the supply—that it will not be interrupted.

Then, you must have plenty of water. The quantity of water used by a big steam turbine plant is something you do not conceive. I know, as every engineer does, the per capita water consumption of a mixed residential and commercial community such as Detroit. I also know the amount of water required to condense the steam for our big turbines. A plant of 100,000 kilowatts, operating at 50 per cent load factor, will require to pass through its condensers approximately the same amount of water as will be used by a mixed residential, industrial and commercial community having a population of 750,000 people. In other words, my power plants take up from the river, heat 10 or 15 degrees, and reject to the river again, more water than is pumped for the whole Detroit community. There-fore, the water supply is the next thing to consider after a good sup-ply of fuel. You must locate on navigable rivers or streams, or on the Great Lakes. Cooling towers are devices you must forswear if you are working for maximum efficiency. One of the appropriate places for a cooling tower is in the dry Southwest where the rapid evaporation in an exceedingly dry atmosphere cools the water sufficiently, and the night chill gives you a fair start the next morning. To locate a big plant on a country duck pond is foolishness.

The next consideration of importance is the labor conditions. Power house labor is changing its character very much. There was a time when the power house got all the incipient geniuses; that was a mistake. But the day has come for brains, and they are being secured. Today the power house employes expect good pay and good environment. These employes will jump in and do a dirty job and do it handsomely, but they are men who appreciate bathing facilities. I recall a tale of an old-time engineer who discussed the showers which were being placed in the plant. He said he took "one of them showers" every Saturday whether he itched or not. But the engineer now gets his shower when he goes off watch, and sometimes he calls his relief and takes a shower while his turn is on. You must supply first-class working conditions, both inside and outside. These men expect their outside, off-work conditions to be such as they can pleasantly live in when they are off watch, where the older men can bring up their families and the younger members can hope to persuade "the only girl" to come.

And, remember, the right kind of labor is a big item. You will have to build, control and sanitate your place, if you locate away from a pleasant community.

If you locate on a place where people consider the outer appearance you must employ an architect. I can think of three power houses where the outside appearance is a credit to the public spirit of the board of directors, while the inside is a libel upon the architect. Foundations will run to $5 per kilowatt when you locate in a marsh with rock 100 or 150 feet down, and you must pile or go down to that rock; or if you locate on the Ohio River where you have 70 feet of rise and fall, there you have an-other serious problem of foundations.

These things balance up. Low fuel cost can be offset by poor water supply. It is better to go a little further away from the coal to get better car service and so on. The next thing is whether you will locate close to your load, or will transmit.

Location versus Transmission

Now, transmission is expensive. It has a very definite measure as to its capacity and a high figure as to its unit cost.

The right of way of your transmission lines is an item of high cost. The line loss may be a high cost. Some interesting figures were obtained when I was asked to furnish the comparative American costs of electric transmission of energy from coal mines against railroad movement of coal, for comparison with possible English costs, with an assumed line loss of 15 per cent. The preliminary result of the comparison is that England cannot move energy by freight train as cheaply as by electric transmission lines, while American freight trains hauling short tons of coal at something like one-third of a cent per mile, can move it more cheaply than the transmission lines. There you see how you must balance one thing against the other.

Again transmissions are not utterly reliable. This was a year notable for lightning, when lightning was painful and frequent and free. We have such things as lightning arresters and they take care of secondary surges; but when you get a real direct stroke of lightning you go around with a shovel and clean up. Let the temperature be 30 degrees below zero, and the electric stress be hundreds of kilowatts; and then change in due course to our hottest summer weather. The porcelain insulators are supposed to adjust themselves to all of these conditions, while placed inside of a rigid metal cap. Well, there are limits to human beings and there are limits to the elements which constitute these strings of porcelain insulators. Thus you will find that lines which behaved splendidly for seven or eight years, after that time behave miserably. And the cost of changing insulator strings is a very considerable item. There are three to a tower and at $10 each makes $30 for each tower, and in a hundred-mile line that is an item. Then do not forget cyclones.

You must balance permissible unreliability against costs. If you are generating by steam power it is probable that your big center power plant will be, per se, more reliable than any plant you can install on your own premises; but the intervening link may change the balancing conditions. You must consider whether you want the continuity of your chemical process to hinge upon the unavoidable conditions which I pointed out as inherent in the transmission line; and you must balance better locations for the power plant against line losses assumed at 15 per cent. That is high. You may spend money and cut it down.

The distribution loss I spoke of earlier is of the order of 20 per cent total. I have got here some figures from a big system: The total distribution loss is 18.7 per cent; of that 5.6 per cent is lost in the line, 3 per cent is lost in the substation transformers; loss in the synchronous condensers—mighty few of them—is one-tenth of one per cent; the conversion loss in changing from alternating to direct cur-rent is 4.7 per cent; storage battery loss is of the same order as synchronous condenser loss; and then there are distribution losses proper, line transformers, meter shunts, and slow meters to make up the difference. It may be considered absurd to put "slow meters" in as loss, but you have the alternative—you can keep them just a little fast or just a little slow; and you will keep them a shade slow when you have lived with them as long as I have.

So far as you are concerned I think the question of location versus transmission turns not on the question of reliability or unreliability, as duplicate lines could alter that; it turns on your own problems of materials and markets. The answer is, if the power cost controls, you must go where you can make power cheaply; other-wise, you must consider your raw materials and markets for your products and then make the best adjustment you can between reliability of transmission, the cheap power location, and your own conditions. It would be possible, I think, and I have said so to some of your members, to choose one of a considerable number of centers where a good adjustment could be made between all these conditions, and then the conditions of plant design would be predicated very closely.

Some Considerations for Future Design

As to generating costs with present practice, 19,000 heat units are entirely practicable. The other operating costs are of the order of one-tenth of a cent per kilowatt-hour. You can figure the cost of your 19,000 heat units, plus one-tenth cent—and you, might note that 17,000 heat units is possible, without any radical departure from present methods or materials.

What is the possibility of higher steam pressures? I speak of pressure above our present 250 to 300 pounds. If you get to consideration of the fact that you are working with high tangential velocities in the turbines, and talking of pressures of 300 to 400 pounds per square inch, and temperatures of 700 Fahrenheit and upward; if you consider that these pressures and velocities and superheats and temperatures are all present—you will realize that the selection and production of suitable materials of construction is a problem. I am told no steel manufacturer will guarantee analysis or qualities of steel at the present time. If that is true the ideal turbine will have for the present to be a thing to be dreamed of and not built. The technical difficulties are solely those of materials. The commercial reasons are likely to keep us under 300 pounds steam pressure and under 700 Fahrenheit, for total steam temperature, for some time to come. The history of the things which already have been tried to get beyond that stage is unending. Some of us are now using high carbon boiler tube because we punish soft steel too much already. The stories of warped castings, and shrink-fits that did not stay shrunk, and erosion of blading, etc., are endless—and time flies.

As to gas fuel : Yes, there is a chance for betterment of economy in gas fuel. You can control your mixture, lose no carbon in your ashes, and have higher heat transmission values. But gas as a commercial proposition, in almost all of the United States, is entirely dependent for its cost upon the by-product market; and that (as you know) is anybody's guess. In the meantime gas producers, as an addition to boilers, cost too much. No combination in present or recent conditions can be made that will allow artificial gas fuel to be used in steam boilers; natural gas is now too valuable; and blast furnace gas has its uses in its own habitat.

The gas engine has its place. At present the large steam turbine has it beaten for general uses. With free furnace gas or natural gas, the gas engine has its commercial merits. But even in the blast furnace business, the steam turbine, with the turbo-blower, has possibilities of displacing the gas-blowing engine. I am told that so long as the intermittent blast of the big blowing engine tends to prevent hanging, they do not want centrifugal blowers. That may be a good practice point, and may be controlling.

Pulverized coal must be considered in any new design. It is first-class now in big locomotives because of the impossibility of firing by hand the present big machines. There is a better control of the air mixture; and you cut off your fuel, and that is the end of it. It may permit the use in power plants of coals which now work very badly in stokers. Please remember that coal is not always coal; it is sometimes an invention of the devil, which should be called by an-other name. Coal that works into a kind of molasses candy is an exasperation, and not a fuel. Coal that fluxes down your furnace walls, and coal that must have acid furnace brick or basic furnace brick—the coal for one kind of fire brick may be the worst coal for another kind—these may be dealt with by pulverization.

Now, the whole question of pulverized coal turns upon the point that it may make coals useful, which with stokers of the present time are not useful, and thereby may reduce total power cost. The big point is to adapt the plant design to the fuel you are going to use.

Now, in closing: It is not good to prophesy, but I know some of you expect to hear from me whether I anticipate any great change in the art. I do not, just now. Dr. Lardner went on record as saying no steamship could carry enough coal to make a voyage across the Atlantic. That was accepted in the thirties, while in the forties there was a steamship line established. Therefore, all I can say is—there is not in sight any great change, unless we can find the materials to take us into much higher temperatures. Boiler and furnace efficiencies are now between 80 and 90 per cent, including economizers, and we need not expect much change there. The remaining losses are commercial losses which cannot be saved profitably; such as the losses necessary to the stack to give you the necessary draft; or are like the big losses of heat thrown overboard, they seem to be inherent in steam. There may be a new cycle produced which will save all this. There is a man waiting for me to come home, ready now to produce that cycle—he says. I hope he has it.

I see also that the freeing of unlimited power is to be investigated by a national appropriation. I do not think that development will be an improvement in the direction we discuss tonight, in the making of electricity by steam power.

I have not told you all I know, but I have told you all I can crowd into a reasonable time, with a bad start.

Home | More Articles | Email: