*Notes on a series of lectures delivered by Dr. C. E. Lucke before the Post-Graduate School of Engineering, and sonic practical applications of these theories.
In spite of the fact that the present types of marine boilers have been in service for a great many years, comparatively little scientific; information has been available regarding the transmission of heat through the heating, surfaces of these boilers and the manner in which such heat transfer could be rendered more efficient. It has generally been conceded that the tubes should be kept clear of soot and dirt accumulations on the gas side and scale and corrosion on the water side; but information regarding improvements in capacity, economic steaming rates and heating surface efficiencies are lacking. Some study has been made of furnaces and methods of firing, together with the intelligent application of the results of flue gas analysis, all of which has resulted in increased boiler efficiency. This is natural since any increase in the furnace efficiency and consequently of heat available for absorption is certain to increase the over-all efficiency; but one point seems to have been overlooked in these investigations, namely, the increase of heating surface efficiency by the use of high gas speeds. If such high gas speeds can find a practical application with high furnace efficiency it is certain that greater boiler efficiencies will result, with a consequent saving of weight and space of incalculable value to a man-of-war.
A number of experiments have been made from time to time to find a, relation between heat made available for absorption and heat absorbed, but without success, until a law was evolved by Mr. H. P. Jordan based on the theoretical assumption made by Prof. Osborn Reynolds, some 50 years ago. It has remained, however, for Dr. C. E. Lucke to show how such laws may be applied to present-day boiler design and operation, and where improvement may be made. The writer here has merely tried to show the line of reasoning followed by Dr. Lucke in his deductions and show how his theoretical assumptions have been corroborated and substantiated by Professor Nicolson in his attempt to improve boiler efficiencies by the use of high gas speeds.
HEATING SURFACES
I. ABSORPTION OF HEAT BY HEATING SURFACES
A metal tube containing water flowing in it and having hot gases passing on the other side results in the formation of a thin film of gases on the gas side of the tube and a thin film of water on the water side of the tube, as illustrated in Fig. 1.
{figure}
FIG. 1
These films of gas and water are the result of friction; that is, they are caused by a rubbing off of particles of gas on the tube on one hand and particles of water on the other hand.
From the Smithsonian Physical Fables it can readily be shown that, taking a value of unity for the resistance of the metal to the flow of heat, the resistance of the water film will be measured as some hundreds while the resistance of the gas film will be some thousands. It is evident, therefore, that o'i the three resistances encountered in the flow of heat from gas through the tube to the water, this gas film offers by far the highest resistance. In these considerations, it is assumed that the tubes under discussion are free from dirt and soot deposits. From this it is evident that the controlling factor lies on the gas side and that to obtain greater heat absorption by the heating surfaces the controlling resistance must be diminished. Practically, this can be accomplished only
by increasing the speed of the gas which, according to Dr. Lucke, rubs off some of the gas film and, according to Professor Nicolson, results in a scouring action along the tubes. In either case, however, the result is to diminish the thickness of this gas film and hence to cut down the resistance offered to the flow of heat from gas to the metal. In like manner, the thickness of the water film
on the water side of the tube, may ,be reduced by increasing the circulation. However, it is evident that, since the controlling resistance lies on the gas side of the tube, no increase in efficiency is to be expected if the circulation alone is increased; that is, without cutting down the thickness of this gas film. This line of reasoning has been amply substantiated by experiment. A series of experiments were carried out, under the direction of Dr. C. E. Lucke, by Mr. W. D. Monks, of Columbia University, showing conclusively that an increase in the speed Of water through the tubes did not increase the rate of heat absorption due to this fact provided the speed of the gas remained constant, thus showing conclusively the existence of this gas film previously referred to.
If, however, the gas film is reduced, it is to be expected that a reduction of the water film by increased circulation will result in increased heating efficiency.
The above statement has been given in the form of a law by Mr. H. P. Jordan as follows " In every case the rate of transmission from hot gases to metal is found to increase with the increase of flow "—and here a most striking relation is demonstrated by an appropriate selection of the prime variable representing the rate of 'flow. When this variable is taken as pounds of air per square foot of area of air passage per second; the B. T. U.'s per hour per square foot per degree difference of temperature between air and metal the relation is linear and the curve is a straight line. These lines have the equation U=A+B (w/a), where U=B.T.U.’s per hour per square foot area per degree difference of temperature between gas and metal. w=pounds of air per second, and a=square feet of cross-Sectional area of gas passage. A =some constant probably dependent upon the cleanliness of the tube. B=some constant dependent upon air temperature and dimensions of air passage, Thus;
B=c+c'q+c"tm,
where c, c' and c" are constants. q, mean hydraulic depth of air passage = also Area of flow of channel/Perimeter of cooling surface, and tm=the arithmetical mean of gas and metal temperatures. These two equations maybe combined as follows:
U=A+(c+c’q+c”tm) w/a,
or, substituting the numerical values found by experiment, this equation as given by Jordan is,
U/3600=.0015 + (.000506 — .00045 X q+ .00000165tm) w/a,
This law may be summarized as follows:
(a) For a constant mass flow; that is, w/a =c, the rate of transfer is proportional to the temperature difference directly.
(b) For a given temperature difference; that is, tm= c, the rate of transfer increases with the speed by linear law.
(c) For a given rate of flow and temperature difference, the rate of transfer increases with the value of the temperature.
(d) The rate of transfer depends on the condition of the surface.
(e) The rate of transfer depends on the size of the channel; and the smaller the ratio of area/perimeter= q, the greater the transfer.
{figure}
FIG. 3
{figure}
FIG. 4
From the above, it is evident that q gives a measure of the ability of the gases to get in contact with the heating surfaces. Thus, if a gas passage is square, as shown in Fig. 3, fewer hot gas particles will come in contact with the heating surface as compared with Fig. 4, which has a greater perimeter but the same area. From this, it is evident that some sort of device for mixing gas currents would be useful for bringing more gases in contact with the hot gases of the tubes; but so far it would seem that no practical devices for accomplishing this have been adopted. The best method would appear to be the use of staggered tubes in the gas passage and having the gas passages so shaped that the stream of gases is broken up instead of flowing in parallel layers, this resulting in a greater number of contacts of hot gas particles with the heating surface in a given length of time. It is also evident that the larger the value of q the greater must be the time of contact of gases or the longer must be the gas passage in order that the same transfer of heat may occur.
2. APPLICATION OF THESE LAWS TO ACTUAL BOILERS
Heat generation is, of course, essential to absorption, but it is necessary to distinguish between an apparent and a real generation; thus, apparent heat generated equals pounds of fuel supplied per hour multiplied by the B. T. U. per pound of fuel as fired.
Real heat gene rated or heat available for absorption is equal to: pounds of fuel supplied per hour x B. T. U.'s per pound of fuel as fired x ( — fraction loss in furnaces).
FURNACE LOSSES
Fuel in the ashes;
Cinders;
Soot discharged through flues;
Unburned hydrocarbons;
Carbon monoxide;
Moistures evaporated.
These furnace losses must be taken into account when the efficiency of the heating surface alone is considered; and in these investigations these furnace losses are charged against furnace efficiency, and remaining, heat in the fuel is considered as heat available for absorption.
The heat available for absorption may be absorbed in two ways, namely, by radiation and by conduction. Radiation is not affected by the dead, gas film and other resistances that oppose the flow of heat by conduction, and therefore does not depend upon the rate of flow of hot gases. According to the law enunciated by Stefan and Boltzmann, radiant heat absorption is proportional to the difference to the fourth powers of the temperatures of the gases and metals. It should therefore be practically independent of the rate of steaming and depend only upon the amount of surface exposed to such radiant heat.
The absorption by conduction follows the laws formulated by Jordan and should therefore be a linear function. The result, according to Professor Nicolson, of combining these two factors of heat absorption Might be given by an equation in the form of x4+bx+c=0. This equation, which might be solved by trial and error, furnishes A solution for the most economical rate of firing for a good design of boiler. The equation as deduced by Professor Nicolson is not given here in full as this leads to considerations outside the scope of this article.
Applying' these two laws to -actual boilers, or to actual boiler tests available, these law S are found to hold good in all cases where the tests have been carried out by skilled experimenters and have been of such a duration as to 'insure accurate results. Plotting B. T. U.’s absorbed per hour per square foot of heating surface
{figure}
FIG. 5
as ordinates, and B. T. U.'s per hour per square foot of heating surface available as abscissa, the boiler test will yield a series of curves as shown by FIG. 5. These curves are Straight lines and have the general equation U=A +b x (heat available for absorption), where b gives the slope of the curve and A is some constant. The method of plotting these curves from boiler tests will be described in detail later. The intercept A on the y-axis is due to the radiant heat as proved by the fact that boilers having the largest amount of heating surface exposed to this radiant heat give the greater intercepts, and vice versa. In the case of the white steam car boiler with no radiant heat the line passes through the origin. The slope of the line, b, would seem to depend upon the mean temperature of the gases and upon the shape of the gas passages; which is in accordance to Jordan's law; that is, it depends upon the values of q and tm. Thus comparing heating surface capacity curves of a boiler under coal with those of the same boiler burning oil, it is noted that the radiant heat intercept is less for the oil than for the coal while the slope is greater for the oil than for the coal. This is to be expected, since tm in the former case is the greater. Other things being equal, the value of q will materially affect the slope of the curve, since this is a measure of the ability of the gases to make contact with the heating surface.
In plotting the curves described above the furnace loss was taken as a constant ratio between the limits of moderate rates of steaming, since in general it would seem that the boilers are designed for their greatest efficiency between the limits of from 20 to 40 pounds of coal per square foot of grate surface per hour. This is in accordance with the law given by Gebhardt for furnace losses. A few numerical values of A and B are given from actual boiler tests:
Boiler | A | B |
Locomotive | 1450 | .66 |
Heine | 615 | .51 |
Stirling | 235 | .82 |
White Stream Car | 0 | .78 |
Hohenstein | 1368 | .61 |
Niclausse | 1760 | .566 |
Hohenstein on Denver | 1090 | .56 |
CAPACITY OF BOILERS
The capacity is taken as B. T. U.'s absorbed per square foot of heating surface per hour, and in accordance with the laws previously enunciated capacity will equal A +b x (the heat available for absorption). The question then arises as to limitations for this capacity curve. A close study of the boilers now in service indicate that five factors limit this capacity curve in practice and prevent us from obtaining very high rates of absorption. These limits may be summarized briefly as follows:
(A) Local failure of circulating system.
(B) Too little water storage.
(C) Excess moistures or priming.
(D) Efficiency too low beyond a certain limit.
(E) Insufficient heat developed due to: First, to low rate of combustion, and, secondly, to a high rate of combustion with small furnace.
The above are the practical limits to capacity of boilers as designed at present, and if these are overcome the limit may be indefinitely extended, since no matter how high the rate of evaporation goes no injury will be done the metal provided one side of the metal is kept wet. This shows that if these limits may be overcome in practice, there are vast possibilities for increased boiler efficiency and higher rates of evaporation than those extant at present. The following table gives an idea of the numerical values of the rated capacity of boilers at present in MC:
| Boiler | Capacity |
Locomotive Boilers, Fire Tube. | Locomotive, Penn. R. R. | 8.55 |
DeGlehn | 9.04 | |
Cross Compound | 9.78 | |
Hanover Compound | 11.88 | |
Simple Consolidated | 12.05-12.39 | |
Baldwin Compound | 14.11 | |
Cole | 16.34 | |
Italian Cruiser | 20.00 | |
Marine Boilers, Water Tube. | Normand | 11.4 |
B. & W. (coal) | 14.6 | |
B. & W. (oil) | 16.0 | |
Express | 10.0 | |
Roberts Coil Boiler | 12.4 | |
Hohenstein | 16.5 | |
Stationary Water Tube. | B. & W. | 10.2 (300 per cent of rated). |
Heine | 4.1 | |
Stirling | 7.2 | |
Cahall | 6.6 | |
Heine | 3.7 | |
Edgemoor | 11.0 (5 inch water draft). | |
Redondo | 6.7 | |
Stationary Fire Tube. | H. R. tube | 6.7 |
H. R. tube | 6.4 stoker fired. | |
Gunboat boiler | 4.0 | |
Lancashire | 11.4 (forced) | |
Bone | 20.0-40.0 |
The tables give the limits commonly used in practice but these limits may be greatly extended provided other things are taken care of.
(A) Local Failure of Circulation.—This is caused by (1) failure of water to get to the spot, and (2) excess heat locally applied.
Among the remedies which suggest themselves for (2) are: Jacketing of tubes exposed to excess heat or subject to radiant heat; this is impracticable, however, since it deprives the heating surface of the effect of valuable radiant heat. Increasing the supply of water to the spot; this may be done by changing to some extent the circulating system to supply a greater flow of water to the tubes exposed to radiant heat. Insure greater discharge of steam and water from the spot subject to excess heat locally applied; this may be done by altering the shape of the discharge end of the tube, provided always, of course, that this does not interfere with the proper circulation of water. Insure water supply to the tube so exposed in such a manner that no other tube robs it of its supply of water. In the B. & W. boiler this might be accomplished by the use of outside down-takes or down-comers. In general, it might be stated that the steam will be discharged most readily from vertical tubes, and least readily from horizontal tubes, increasing with the increased slope of the tubes to the horizontal. When seeking to increase boiler capacity it is necessary to insure an active circulation and the ready access of water to parts subject to excess local heating together with an easy discharge of steam and water from that tube. The header type of boiler is poorly designed to meet these requirements. Scale will naturally Cause the failure in any boiler; hence, when designing for high capacity boilers must be so designed as to permit of ready access for cleaning and for the prevention of scale in the tubes.
(B) Too Little Water Storage.—(1) Too little total water. (2) Too little drum water.
Too little total water causes the boiler to be too sensitive to fluctuations, while, on the other hand, too much total water makes the boiler slow to respond to sudden demands for extra steam. The time to evaporate the total water at the designed moderate steaming rates must be taken into account; that is, the time taken to evaporate all the water in the boiler with the feed shut off. Too little drum water will cause surging, and, consequently, the danger of the water level dropping too low. Hence, there must be a sufficient storage of water in the drum to prevent such accidents. When designing for high capacities these points must be carefully considered.
(C) Moisture.
Boiler | Equivalent evaporation | Percentage of moisture |
B. & W. under coal | 4 | 0 |
B. & W. | 15 | .43 |
B. & W. under oil | 4 | .11 |
B. & W. under oil | 16 | .81 |
Normand | 4 | .14 |
Normand | 9.75 | .41 |
From the above it will be seen that in general the moisture increases with the steaming rate, but there is a certain range of steaming rates where the moisture is inappreciable. The practical danger limit for moisture depends upon conditions, but, in general, 1 per cent is the limit. When using high rates of evaporation some control of this moisture is necessary, as it is essential that the machinery be supplied with dry steam.
CONTROL OF MOISTURE
Moisture will depend upon the disengaging surface which controls the velocity of escape of the steam bubbles. Steam bubbles escaping into the steam space at high velocities or at what is called a high disengaging, velocity will tend to spatter water all over the steam space and cause such particles of water to be carried off with the steam. Several preventatives suggest themselves as a practical remedy for this priming.
(1) Baffles in the water in order to reduce the velocity of steam bubbles passing through the water. This remedy might be used in boilers where, the steam and water are discharged below the water level.
(2) Baffles in the steam space; that is, splash plates for boilers discharging in the steam space above the water level.
(3) Increased steam space.
(4) Extra drums.
(5) Dry pipes.
In general, the disengaging velocity, is inversely proportional to the disengaging surface. Thus, a Scotch boiler having as a disengaging surface the entire surface of the water in the boiler should have under normal Circumstances a low disengaging velocity, and the same should be the case for vertical water-tube boilers. Boilers having a small steam drum and, consequently, a smaller disengaging, surface, will have a higher, disengaging velocity and will, therefore, tend to prime unless some contrivance is adopted to prevent this carrying over of moisture. Another manner of reducing the disengaging velocity is to use increased steam pressures, thus increasing the density of the steam and decreasing the disengaging velocity, since the disengaging velocity= cu. ft. per hour/ disengaging area. In the latter case the specific volume of the steam will be reduced, and a low disengaging velocity will result. The use of high pressure with reducing valve might, therefore, be advantageous under some circumstances, since the wire drawing through the reducing valve tends to superheat the steam and insure a Supply of dry steam to the engines. The disadvantage lies in the increased cost and weight of material necessary to withstand high pressures.
(D) Loss of Efficiency with High Capacity.—The absolute limit here is given by the point where any increase in fuel fired results in no increase in the amount of steam formed. This is termed the economic limit of the boiler. It will be seen from considerations taken up later that this limit is set by the furnace alone and that increased efficiency of this factor would result directly from the increase in the furnace efficiency. The above is evident, since the heating surface will absorb a fixed percentage of all of the heat brought in contact with it, other things being equal.
(E): Insufficient Heat Developed.—The amount of fuel burned per square foot of grate surface per hour does not necessarily depend upon the draft alone, since certain cases might arise where draft was very high but a relatively small amount of heat is being generated due to improper thickness of the fires. On the other hand, fires might be of a proper thickness but the manner of admitting air to the furnace might result in incomplete combustion of the fuel. Thus it will be seen that the manner of firing and the proper thickness of carrying fires for different rates of steaming and conditions: of draft are factors which must be seriously considered when attempting to obtain high efficiencies in any type of boiler. Officers desiring to make a study of the proper methods of carrying fires and burning the coal on the grates, together with the considerations of sizes of furnaces for good combustion of the fuel and high efficiencies, are referred to the most excellent paper recently issued by the Bureau of Mines, entitled "Factors Governing the Combustion of Coal in Boiler Furnaces."
{graph}
PLATE I.—II. S. capacity curves N.B.—B. T. U.'s available= fuel beat X constant % furnace for moderate rates. 5% coal. 2% oil fuel B. T. U. absorbed per a H. S. per hour boilers under oil. Tested by the Fuel Oil Testing Plant boilers under coal. Melville Tests
Taking the data from the series of boiler tests a set of curves may be plotted as illustrated in Plate 1. These curves contain no additional data over those of the test from which they were taken, but when plotted in this manner, illustrate clearly the losses occurring in a boiler. They are instructive in showing how over-all efficiencies may be improved by improving the efficiencies of the heating surfaces and furnaces. It is evident from a perusal of the diagram that the heating surface efficiency will depend very largely upon the slope of the heating surface capacity curves, and we know from Jordan's law that the slope of this curve may be increased by increasing the value of tm which is the mean temperature differenee between gas and metal or, in other words, the temperatures of the flue gases, and also by decreasing the value of q which is the. mean hydraulic depth of the gas passage. From this it is evident that high gas speeds with suitably designed areas for gas passage must inevitably result in high heating surface efficiencies and, consequently, in a greater over-all efficiency of the boiler itself. The heating surface efficiency curve is seen to slope down as the capacity increases and approaches, but never crosses, the line marked b to which it is asymptotic, b being the slope of the curve. (Plate 4.) This has been proved to be the case in a series of experiments carried out on an experimental boiler by Professor Nicolson, in which the gas speeds were very high and resulted, according to him, in an average heat transmission of 46,400 B. T. U.'s per square foot per hour, corresponding to an evaporation of about 48 standard units per square foot per hour. This result is very much larger than that usually obtained in practice, but was subsequently confirmed by Mr. Longridge's test on the same boiler. It is interesting to note, also, in connection with these tests that while the temperature of the gases in the combustion chamber was 2600° F. the temperature of the gases after leaving the economizer or the smoke-stack temperature was only 279° F., showing that a relatively large percentage of the heat was absorbed by the heating surface, since the smoke-stack temperature was lower than that usually found in practice.
The boiler used by Professor Nicolson in carrying out his tests was a Cornish boiler arranged as follows:
In the last 10 feet of boiler flue there was inserted a circular brick plug 38 inches in diameter and 10 feet long, having an annular space 1 1/2 inches around it for the flow of gases. The gases then passed into an evaporator—passing along the tubes through which water circulated. The evaporator (16 inches in diameter) contained 90 tubes of 5/8 -inch outside diameter, and the center was filled with a pipe 6 inches in diameter to which the gases had no access. The gases were then forced through an economizer 16 inches in diameter Containing 163 vertical tubes of 7/8-inch outside diameter.
The gases were drawn 'through the economizer and evaporator by high speed fan 1640 r. p. m.
Thus the areas for gas passage were as follows: Flue 186 square inches. Evaporator 145.5 square inches: Economizer 103 square inches.
In a subsequent arrangement for the tests conducted by Mr. Longridge these dimensions were changed slightly, but ' the resultant gas speeds were approximately as follows: Flue 79-152 foot-seconds, evaporator 98-152 foot-seconds, economizer 83-152.
Thus the gas speeds were increased to about three times those used in present-day practice, reducing the gas film considerably. The water film was reduced by increased circulation in the evaporator and economizer tubes by means of square rods inserted in the tubes, thereby reducing the area for passage of water and increasing the rate of flow of water through them.
On Plate 2 are plotted the results of these tests—showing that when boiler, evaporator and economizer were taken separately the absorption followed the straight line law of Jordan. When plotted altogether there is:a slight deviation from the straight line law due to the varying values of q or mean hydraulic depth of gas passage in boiler flue; economizer and evaporator, giving each line a 'slightly different 'Slope when plotted independently. The result of 'a test on a good locomotive boiler quoted above is also plotted on this plate for purposes of comparison.
In plotting these curves the furnace losses were taken according to Mr. Longridge's report at about 5 per cent—which is thought to be rather conservative for a boiler using coal. If the furnace losses were greater the slope of the lines would be even further increased, giving a still larger heating surface efficiency.
The heating surface efficiency curves and boiler efficiency curves are shown plotted on Plate 3, being plotted directly from the carves on Plate 2; and Mr. Longridge's tests. These curves indicate that the furnace efficiency was probably assumed a trifle too great, since the over-all efficiency falls off somewhat at higher rates, due to falling off furnace efficiency. This is evident since the over-all efficiency is the product of heating surface and boiler efficiency curves.
{figure}
Plate 2
N.13.—Furnace losses as per data from report of tests. Average 5%
The conclusions drawn from these experiments are:
(a) High gas speeds increase the slope of the heating surface efficiency curves.
(b) Much greater boiler capacities are possible than those now used in practice.
(c) Greater over-all efficiency is possible with better designed furnaces.
{figure}
Plate 3
(d) Economizers with their subsequent saving are feasible for naval use when used with high gas speeds.
If a practical boiler can be constructed embodying these principles its advantages to the service would be enormous.
In these tests above quoted Professor Nicolson made use of economizers in the last gas pass in order that all the heat possible might be extracted from the gases. And it is interesting to note here that while economizers at present designed for moderate gas speeds are too large and unwieldy to justify their use in the naval service, yet with high gas speeds a greater heat transmission may
{figure}
PLATE 4
B. T. U.'s absorbed per ?' H. S. per hour
N. B. —Boiler efficiency = H. S. eff. X furnace efficiency
These curves were plotted from curves on Plate I and boiler efficiency data from tests
Heating Surface Efficiency Curves
be effected and the size of these economizers materially reduced, thus permitting their adoption in the service with the consequent saving and gain in efficiency.
In a subsequent test of this same boiler and slightly different arrangement of the heating surfaces, Mr. Longridge obtained a net boiler efficiency in one case as high as 79.1 per cent, and deducting therefrom the steam for the 'fan the percentage of efficiency was still 68.9 per cent, showing that the gain in efficiency more than overbalanced the loss of steam slue to running blowers at high speed. As a result of these tests Professor Nicolson has stated that: “It may confidently be expected that a high speed counter-flow boiler to give a net efficiency of 80 per cent when evaporating at a rate of 20 pounds of steam per square foot of heating surface per hour will shortly be constructed."
With such high gas speeds it must be remembered, however that the furnace efficiency must be taken care of, both in design and in methods of firing, in order that a great falling off in efficiency may not result; and also it is essential that the five limits to heating surface capacity previously discussed should not be lost sight of.
Professor Nicolson, in a suggested design for boilers embodying the principles laid down above, Claims that the following advantages to men-of-war may be expected: “(1) Saving of boiler space of 30 per cent; (2) Saving in boiler weights of 30 per cent; (3) Economy of coal consumption of at least 5 per cent; (4) No flame at the funnels even at full power; (5) No smoke; (6) Funnels may be of any height sufficient to carry gas over bridges, and need only have 6o per cent of the present area; (7) Steam may be got up from coal in 15 minutes, provided, of course, there is standby steam available for the fans; (8) Great handiness for changing power by regulation of fan speed; (9) No closed stoke holes or ash-pits."
The writer has purposely omitted a detailed discussion of the proposed innovations of Professor Nicolson, since they are all incorporated in a paper read by Professor Nicolson before the Institution of Engineers and Shipbuilders in Scotland, and since the object of this paper is simply to call the attention of the service in general to the possibility of increased efficiency in boiler design and management. A systematic study of these ideas, as enunciated by Dr. Lucke in his lectures, should he invaluable to engineer officers desiring greater efficiency in the plants under their charge.