REPORT

On an overseas business trip.

In the first days of August 1905 I went abroad and stayed there until June 1906, that is, a total of 10 months.

My work during this time consisted of the following: first, becoming acquainted with the teaching of turbine construction at the Zurich Polytechnic; second, inspecting hydroelectric stations already in operation or under construction and, finally, visiting turbine-construction works.

Since I had to deal with all this without any definite system, in my opinion it will be much more convenient in the present report to set forth the material by individual questions, and not in chronological order; and in view of the fact that, in the course of time, certain questions having a direct connection with turbine construction I intend to examine in greater detail, I present them in a rather compressed form, dwelling most of all on those subjects to which I do not think I shall ever return. Therefore this work is somewhat uneven in its presentation, for which I ask the reader’s indulgence in advance.

The teaching of turbine construction at the Zurich Polytechnic.

The teaching of hydraulics at the Polytechnic is divided into two departments: general hydraulics, which is read for one semester by Prof. Fliegner, and hydraulic engines and pumps, read by Prof. Prasil.

Prof. Prasil’s lectures continue over three semesters, beginning with the second year. First, “Hydraulische motoren und Pumpen” is read (one semester, 2 hours per week), a course covering piston pumps and the beginning of turbines (axial reaction turbines).

In the following semester this course is continued (4 hours per week); here I note the chapters on Francis turbines of the high-speed type, Pelton wheels, high-pressure centrifugal pumps, and automatic regulation of turbines, the treatment of which is entirely absent from our higher schools.

In addition to the compulsory courses set forth above, there is also a small supplementary course for hydraulic engineers (one semester, 2 hours per week), “Ausgewälte Kapietel über hydraulische Anlagen.” Here are presented the theory of water hammer in closed conduits and the hydrodynamic theory of Francis turbines, of which Prasil himself is considered the creator. This course, however, is rather abstruse and difficult to follow, but, as I have already said, it is not compulsory.

In parallel with the lectures in the course, classes are held in the hydraulic laboratory. These classes consist chiefly in determining the operating diagrams of turbines of various types and under various conditions; in doing so, students must become acquainted in practice with all sorts of measuring apparatus and instruments, such as Amsler wings, revolution counters, water gauges, Prony brakes, etc. The observations made are entered on specially ruled sheets, and each of the participants is given one copy of the record for preparing the diagram and report at home.

For work in the laboratory the students are divided into groups of 8–10 persons each. Each group works 4–5 times during the semester; the classes are conducted under the direction of a laboratory assistant, while the professor looks in only occasionally for a few minutes.

Special attention is devoted to the turbine design project. This project takes up the whole winter semester, i.e. about 20 weeks, with the possibility of working in the drafting room on all weekdays from 12 to 7 o’clock in the evening.

The majority of students are assigned Francis turbines of the high-speed type; to some few, who finish the work earlier, a Pelton wheel is assigned, and, if they wish, they are also invited to design a centrifugal pump.

The time devoted by the teaching staff to this project is very considerable. Thus, in my time, for 120 persons doing the design work there were 1 professor and 4 assistants; all these five persons were present 4 times a week for 3 hours, which amounts to 60 hours per week, or half an hour of time weekly for each student. Such an abundance of time allotted to the students, of course, does not fail to affect the quality of their projects, which, generally speaking, are considerably better than the projects in our institutes.

However, not all students complete their hydraulic education with this project. Among the diploma projects there appears: “Wasser-Kraft-Zentralanlagen.” The project is usually connected with the electrical transmission of power and is assigned at the same time to electrical engineers as well. Both they and the others must present considerations on the project as a whole, but detailed calculations and designs only in their own specialty.

In order to arouse greater interest, the project is assigned on the basis of existing investigations, with an indication of the locality where it may be carried out.

During my stay the diploma project assigned was: “Das Albulawerk der stadt Zürich.” Its history is as follows:

For several years already the city of Zurich had been concerned with obtaining cheap energy, since the existing station on the Limmat River was proving insufficient. In order to resolve this question, a competition was announced for the preparation of a project. Last year the time for submitting projects expired. Three projects were submitted, which were examined by a special commission that adopted the construction of a hydroelectric station on the Albula River, with transmission of energy of 24,000 horsepower over 135 kilometers, at a voltage according to one variant of about 50,000 volts (with three-phase currents) and according to another of about 80,000 volts (with direct current, series system).

After accepting the project, the city allocated to this enterprise, which is to be completed in 1909, about 10.5 million francs. (According to the estimate, about 5 million for the hydraulic part and 5.5 for the electrical part.)

All the competing designs were put on display in the city hall throughout the spring for the public to inspect; special days were appointed for the students, on which explanations were given to them. After all this, the designs were also made available as a diploma assignment, taken up by many students. Each was invited, at his own discretion, to modify and alter them as he deemed best; the main task, however, consisted in the detailed development of the machine designs.

Concluding my note on the Polytechnic, I cannot fail to point out the outstanding role, in the matter of technical education, of frequent excursions, which are undertaken both to the sites of already existing stations and to places where such stations may be built, as well as to turbine-manufacturing works.

Exploitation of Water Power in Switzerland.

If one takes into account the small area occupied by Switzerland, then in the exploitation of water power this country appears to be first in the world. A large share in this is due to improvements in the transmission of power by electricity. Here, in the first place, stands the effort to transmit currents of ever-increasing voltage. Whereas fifteen years ago the first large hydroelectric stations were built for transmitting currents at a voltage of 3–5 thousand volts, now they already go up to 40,000 volts with alternating current; and according to the design prepared for the station on the Albula River, in one of the variants of transmission, the voltage with transmitted direct current reaches even as much as 80,000 volts.

The technical possibility, increasing with each year, of transmitting currents at ever higher voltages—whose limit is apparently not bounded—entails an increase in the distance over which power can be transmitted and makes profitable those enterprises which, only a few years ago, did not pay for themselves.

According to statistical data from 1888, i.e. from the time when small hydroelectric stations had already appeared, within Switzerland 110,000 water horsepower were being exploited. In 1902, 160,000 horsepower went to the production of electricity, and 90,000 horsepower to other purposes. In 1905 the total energy produced by water turbines was estimated at 300,000 horsepower, which almost entirely went to electricity.

According to more precise information from 1902, the electrical energy obtained from water was distributed as follows:

Under what diverse conditions hydroelectric stations were built can best be judged from the selling prices for electrical energy that now exist in various places of the republic. According to Pauschel: energy for 10 incandescent lamps of 16 candlepower each, at 1,700 annual hours, is sold for 128–750 francs; for 100 such lamps at 500 annual hours, 800–2,500 francs; for supplying a motor of 3 HP at 3,000 annual hours, 100–325 francs; for supplying a motor of 5 HP under the same conditions, 135–420 francs.

According to the project on the Albula River, the cost of energy in high-voltage current within the city of Chur will be, for itself, 85 francs per kilowatt-year at 11 hours daily.

The sale price of energy for motors is usually considerably lower than for lighting. The reason for this lies in the desire of entrepreneurs to attract more daytime consumers, i.e., at times when there is usually a surplus of energy.

The considerable difference in the cost of electrical energy is at present a subject disturbing certain Swiss patriots; through the press they are trying to popularize the idea of the necessity of buying out the existing hydroelectric stations from private hands, as well as presenting to the government a monopoly on the construction of new ones; under these conditions, in their opinion, the price of energy for all localities of the republic could be made uniform, which in turn could greatly raise the entire industry of the country.

This is not the place to examine this opinion; I shall note only that the shareholders of the present enterprises are content with a very small profit, although one that is growing from year to year owing to the increase in the number of consumers; until now, however, at not a single large enterprise does it, after deduction of amortization, exceed 6% per annum, while on average it amounts to only 4%.

All hydro-power stations in Switzerland may be divided into two types. The first includes installations with low water pressure, up to 12 meters. The second includes those with a higher pressure, from 50 to 500 meters. The absence of installations between 12 and 50 meters of head appears somewhat strange; an effort is made to explain this strangeness, with some strain, by the structure of the soil; it is more likely, however, that this is an accident which may disappear with time.

The majority of the large stations inspected by me have always had a further small steam reserve installation, consisting either of a steam engine connected directly with a dynamo or, in the newest constructions, of a steam turbine with a dynamo. Some of these reserves operate for only a few days a year, and rather so that the machine should not stand idle than from necessity; nevertheless they are necessary in view of possible contingencies. Of course, at small stations they manage without a reserve installation, but at first-class stations its presence is almost the general rule. According to the observations I made, in those cases when there is, in general, sufficient water power, the reserve for the most part amounts to 25% of the available water power.

In those cases when we are dealing with a low-pressure station, as a reserve there is sometimes used an arrangement whose significance will become clear from the description of the central hydroelectric station “Olten-Aarburg.” Here the low-pressure turbines are situated on the Aare River. During the hours of greatest energy consumption for lighting, there is a shortage of it, while during the day there is a surplus. To take advantage of this surplus, the following arrangement was devised: on a hill, 300 meters above the water level in the Aare, a reservoir with a capacity of 10,000 cubic meters was built; in a special building at the foot of this hill there were placed in a row a high-pressure centrifugal pump, an electric motor, and a Pelton wheel; all three machines were installed so that their axes of rotation lie on one line and, by means of couplings, can ...

to fluctuate. During hours of surplus energy, the current generated by the low-voltage machines is transmitted to an electric motor, which turns a pump and drives water into the reservoir on the hill; and during a shortage of energy the pump is disconnected from the electric motor, which is coupled to a Pelton wheel. The water from the reservoir moves back through the same pipe through which it was pumped up, and now makes the Pelton wheel work, turning the electric motor, which now operates as a dynamo and transmits current into the line.

A still older installation of this kind exists at the municipal station on the Limmat, in the city of Zurich, but here the pump is a piston pump.

Some of the low-pressure stations operate continuously with an excess of pressure, for which reason the regulation of the turbines at them is often built with great shortcomings in respect of water economy, but on the other hand is very simple and reliable in construction.

With regard to the types of turbines, it may be said that in the oldest installations the Jonval types with several outlets predominate, whereas now, without exception, Francis turbines of a high-speed type are being installed everywhere, often of double, triple, and even quadruple construction (i.e. 2–3–4 on one shaft).

High-pressure stations present great variety both in the application of engine systems and in the entire arrangement scheme. A necessary appurtenance of most of them is earth and stone high dams, a water-supply tunnel, and long iron or cast-iron pipes. In old installations, Girard turbines were used predominantly, in the Schwartzkopff system, now the Pelton wheel. Bucket wheels (Löffelrad) and the Francis turbine.

Since at the present time many German technical journals have, almost continuously for several years now, been publishing descriptions of many individual hydro-stations, I see no point in describing them here. I consider it more appropriate to dwell on certain special questions.

The first question, the ignoring of which afflicts some of the earliest structures, is the question of the presence in the water of foreign suspended bodies. It is clear that every, even the oldest, structure always contains a device preventing large floating bodies—such as, for example, pieces of wood, branches, and so on—from entering the turbine chambers. This device usually consists of two grates, of which the first, made of thick iron rods spaced at a distance of 100–200 m/m, stands in front, while the second, made of strip iron set on edge, with a gap between the strips of 10–25 m/m, is at the turbine chambers themselves. The latter grate greatly constricts the clear opening, and therefore at the place where it is installed the canal is usually widened, and the grate itself is set at an inclination, which not only increases the clear opening but also facilitates cleaning it with iron rakes from the objects that have drifted onto it. Such a method of cleaning is practiced everywhere and fully achieves its purpose under ordinary conditions. However, there are sometimes cases when so many floating bodies appear in the river that the grate becomes clogged very quickly and such cleaning becomes very difficult. An illustration of what has been said may be the case in January 1905 at Rheinfelden, where for two days the grate had to be cleaned continuously from a drift of small pieces of ice appearing in the Rhine (in our country this phenomenon is called “sa-

“with a crowbar”) 250 workers. Despite such colossal labor, the station’s productivity, owing to the insufficient inflow of water, was considerably reduced.

To facilitate the cleaning of the grating, the usual procedure is as follows: let, in drawing 1, a canal be made in plan, conducting the water to the turbines; then the portion of the grating is placed obliquely to the current, and, at the sharp angle between the grating and the bank, above the grating, a discharge sluice is made (in the drawing, by mistake, the sluice is shown at the left bank, whereas in fact it should be made at the right), by opening which all the floating matter that has come up to the grating can be flushed away; it is understood that the sluice must be of suitable dimensions.

If the floating material is held chiefly at the surface, then, in order to divert it into the sluice, in front of the grating, obliquely to the current, a floating wall is installed, consisting of several wooden beams; along it the floating matter slides easily and is carried away through the sluice, into which the water needed for the given case can be admitted by a spillway.

If, on the contrary, the water is rich in matter moving near the bottom, then a sill is made in front of the grating (see the drawing); this sill holds back the deposits and directs them into the sluice opening.

All these means serve as a good aid for cleaning the grating, but they are never, by themselves, radical, since the silting of the grating also depends on the improper position of the canal feeding the turbines in relation to the entire channel. At Rheinfelden there are: a discharge sluice, a sill, and a floating wall; it is true that the discharge sluice is of very small dimensions, but the chief error lies in the position of the feeding canal. Owing to this very position, the canal, at its entrance, is often silted with sand, so that there is even a special dredge which works every year to clear the deposits.

To clarify what the error is here, let us take (drawing 2) the diagram representing the installation at Rheinfelden. From this drawing it is evident that the entrance into the headrace canal begins in the middle of the river channel, while the dam itself stands at an obtuse angle to the canal. It may be expected in advance that, at low water, when, consequently, water does not pass over the dam, all floating matter enters the canal. The same will happen at mean water, since then, nevertheless, the chief mass of water moves toward the headrace canal, carrying with it a great quantity of floating matter. Now let us take very high water; at this time the Rhine carries a mass of silt and sand; the chief mass of water rushes through the opened sluices of the dam, and the greatest velocity in the riverbed is transferred to the left bank; in the canal, however, the velocity changes little, and it even decreases in comparison with the former: since the level in the canal rises, while the discharge through the turbines, owing to the backwater downstream, remains the same. Therefore the water, entering the canal, considerably reduces its velocity. It is clear that, in this case, sand and silt are deposited intensively; when the water falls, a riffle forms at the place outlined by the dotted line, hindering the entrance of water into the canal. The case described at Rheinfelden deserves full attention on the part of Russian engineers, since many of our rivers, in the character of their flow, are similar to the Rhine.

In the designs of several new structures on the Rhine (Beitrag zur Hydagraphie des grosherzogstum Baden, 12 Heft), a scheme, shown in drawing 3, has been adopted with slight modifications; it seems to me very successful; here, by opening the sluice in the dam, it is easy to achieve proper cleaning of the grating. One should fear

the drift of sand into the head of the canal, even if the turbines are located not so close to the main channel as is shown in the drawing, has no basis; rather, one may expect the opposite phenomenon, i.e. the washing out of the entrance into the canal during energetic movement along the channel and through the dam; but if the head of the canal is paved or faced with stone, then this danger also disappears.

Sand and silt carried by the water are a very unpleasant phenomenon. In low-pressure installations, besides the drift of sand into the canal, sand and silt also often settle on the turbines themselves, disturbing the regulation, wearing away the blades and bearings, etc.; however, with proper design and careful maintenance, this evil is to a considerable degree removable.

In high-pressure installations the harm from silt and sand increases considerably: water, rushing at a speed of several tens of meters through the turbine canal, carries along silt and grains of sand, which at such a speed, possessing considerable living force and striking the blade, wear it down like a grindstone; this wear may proceed so rapidly that an entire blade can sometimes be completely destroyed within several days. The use of more resistant materials for manufacturing blades, such as, for example, steel and phosphor bronze, reduces the wear. The best way of eliminating this inconvenience, however, may be the construction of so-called settling basins. In many installations of this type there are large settling reservoirs through which the water diverted from the river must pass before reaching the turbine. In these reservoirs the velocity is very small—practically equal to zero—and here, consequently, almost complete deposition of sand and silt takes place.

Of more economic than technical interest is the construction in dams of a permanent passage for fish (such a passage the Germans call a “fish ladder”). These passages are arranged chiefly for two kinds of fish, for salmon and trout, which freely overcome a water velocity of 4–5 meters per 1″. The arrangement of such a passage is shown in plan and section (see drawing 4); here the principal dimensions are also shown, in meters, on the condition that large fish can pass.

Among the special features of high-pressure power hydrostations is the conveyance of water by tunnels. Tunnels for such a purpose have small cross-sections of from 3–8 sq. meters, but often reach considerable length: sometimes up to 10 kilometers. Tunnels with a cross-sectional area of less than 2.5 meters are used very rarely because of the difficulty of working in them. In loose soils, the driving of a tunnel is extremely difficult and at times completely impossible. In rock, the entire excavation has to be carried out with dynamite (in Switzerland an exclusively used composition is “Sprenggelatyne,” manufactured by the firm “Isleten bei Flüelen”).

In clayey soils one has to work with a crowbar, pick, and shovel; if the clay has compacted, then work in it is very difficult: in this case the clay has almost the hardness of stone, with the difference, however, that working with dynamite is impossible here, since an explosion causes destruction only at the surface; most often only a small funnel-shaped depression is obtained around the dynamite charge. At the same time, clay readily causes collapses, so that the entire excavation just made must immediately be supported by a wooden structure, and

only then can the masonry lining be carried out beneath it with brick or concrete. Under such conditions, the initial cross-sectional area of the excavation for the tunnel must be made considerably larger than the final one. In view of the foregoing, the cost of a tunnel in clayey soil is almost no less than in rock. Water-supply tunnels are almost always, without exception, lined inside with brick or concrete. Here, besides considerations of strength, an important role is also played by the reduction of resistance to the movement of water, which in a long tunnel is quite considerable.

To make the work cheaper, and chiefly to speed it up, work is carried on simultaneously at several points, and is always begun from both ends; moreover, long tunnels are divided into sections, so as to obtain several exits to the surface at the same time, and for this purpose it is sometimes necessary to replace the straight line of the tunnel axis by a certain broken line with obtuse angles. Although this lengthens the tunnel line, the advantages compensate for the cost of the lengthening. After the construction of the separate sections is completed, the exits to the surface are closed up with masonry or concrete; in this way one continuous tunnel is obtained. Sometimes there are also passages from one section of the tunnel to another by means of iron pipes.

As for the motion of water in the tunnel, it takes place for the most part by gravity or under slight pressure; only in exceptional tunnels does the pressure reach several atmospheres. When working by blasting, from 6–12 holes must be made in the tunnel, arranged in several horizontal rows (see Fig. 5). These holes are chiselled to a depth of 0.6–1.2 meters and to a diameter of about 40 m/m; for chiselling, a rod of round steel is used (Fig. 6), one end of which is forged flat and sharpened. In hard rock the chiselling is done by constantly turning the rod and striking the other end with a hammer. If the rock is soft, then chiselling is replaced by drilling with the aid of a mechanism very similar to a mechanic’s ratchet for drilling holes in metal.

As far as I could observe, the steel used for the rods and drills was exclusively of the “Böhler–Sfeurmark Osterreich” brand. The first row of holes is drilled with a certain upward inclination, the second row—horizontally, and the third—at a downward angle. Pieces of dynamite gelatin, prepared in the form of cylinders, are placed in the holes, several pieces in each hole; into the upper one of these a powder igniter tube with a piston is inserted; the remaining part of the hole is filled with sand, plastered with clay, or sometimes even simply filled with water; then the powder tubes are ignited, the workers leave, and the explosion takes place.

The upper row of holes is exploded first, whereby wedge-shaped stones are split off; then the middle row is exploded and, finally, the lower one. The first explosion produces little destruction; mainly it prepares the place for the action of the following ones.

How many holes should be made, of what depth, and how they should be arranged can be determined for each rock only by experience. In hard rocks, and at the same time those difficult to split, 12 holes are made in 4 rows, and sometimes even 16.

Work in the tunnel proceeds in 3 shifts of 8 hours each; in a shift 4 men work, not counting the haulers and the people for pumping out water, if such is necessary—

that is, it cannot work for more than one month in succession. Under such conditions, on average, they advance by 1.5 meters per day; in exceptionally favorable rock, by 2.5–4 meters. But if two shifts of 12 hours each work, then the productivity per day falls to ¾. In clayey soil the performance of 2 or 3 shifts differs little, since the chief delay here comes from fastening the excavation with wooden structures; in general, in such soil, in a tunnel of moderate depth, it is rarely possible to advance, on average, more than 2–2.5 meters per day. The prices for the work, according to the data I have collected, are as follows:

For a tunnel of medium length (from 1 to 3 kilometers), the cost of excavating one cubic meter is from 15 to 35 francs. The consumption of dynamite per 1 cubic meter of rock is from 1 to 2 kilograms. Workers (miners) receive, for 8 hours of labor, 4–5 francs with their own board; the haulers of the excavated material (usually transported on carts along laid rails) receive 3–4 francs. Two haulers move a wagon with a load of 0.5–1.2 cubic meters. The lining and facing of the tunnel, per cubic meter of facing material, costs 30–40 francs. With such prices, one running meter of a tunnel, with an opening of 4–6 square meters in the clear, costs from 200 to 400 francs. Tunnels more than 100 meters long, and where work is done with dynamite, must necessarily be ventilated; ventilation is carried out by forcing air through an iron pipe, the end of which terminates almost at the very place of the explosion. Lighting is provided by oil lamps. Electricity is not used either for lighting or for ignition, because during explosions the wires are easily broken and become entangled.

Machine work in the construction of water-conveying tunnels is used very rarely, since it is 20%–30% more expensive than manual work and provides only a gain in time.

The pipes used for conveying water to turbines are of reinforced concrete, iron, and cast iron. Reinforced-concrete pipes are of considerable diameter, but in them, as a rule, only slight internal pressure is allowed, which is why they lie almost horizontally and are used comparatively rarely; on the contrary, concrete and reinforced-concrete pipes are being used more and more often for water discharge and as suction pipes. Cast-iron pipes are generally used for small diameters (about 400 millimeters); for larger diameters it is more advantageous to use iron riveted pipes.

The largest diameter of such a pipe that I have seen is 2 meters. (In Sweden, iron pipes up to 4 meters in diameter are used.) The pressure in their lower part may reach large values—in some installations up to 50 atmospheres.

In order to economize material, the pipes are usually divided into several sections, the thickness of the sheet iron in them being unequal: in the upper section it is made several m/m less than in the lower one; they are ordinarily designed with a fivefold factor of safety, but their thickness is never taken as less than 5 m/m. As for the laying of pipes, they are placed on stone foundation piers; the distance between piers is 5–6 meters; sometimes a wooden sleeper, pre-shaped to the proper curvature, is first laid on the stone foundation and fastened to the foundation with bolts, and the pipe is then laid upon it. At points where the pipe turns, where the appearance of forces acting along the axis may be expected,

the foundation piers are replaced by concrete abutment masses having a resistance to weight sufficient for this purpose; on such a mass the pipe ring is fixed by means of special shaped parts and bolts. The harm from the lengthening and shortening of pipes, owing to changes of temperature, is eliminated by the use of stuffing boxes, which are necessarily placed between two rings; moreover, for 100–150 meters of pipe length one stuffing box is quite sufficient. Achieving tightness in the stuffing box under high pressure is rather difficult.

At the Escher & Wyss works, for this purpose, they use the device shown in drawing 7. A cylinder of red copper 4–5 m/m thick, inserted between the walls of two pipes after the inner pipe has been pushed in, is wedged out at the places c and d (see drawing); in this way a very tight joint is obtained, but at the same time one capable of displacement, since the cylindrical surface adjacent to the copper is carefully turned and polished. Tightness between the flanges of the pipes is achieved by laying rubber rings.

In general, any long closed water conduit, whether a tunnel or a pipe, presents a great danger in the event of a sudden stoppage of the flow. In this case the entire mass of stopped water, possessing an enormous reserve of energy, raises the pressure in the lower part of the pipe to dangerous limits. If the closing of the pipe is carried out slowly, as, for example, happens when hand mechanisms are used, then the whole action is weakened to such an extent that no complications are caused. The question of shocks in pipes and tunnels is poorly developed in theory; nevertheless, the complications caused by them are completely eliminated by the arrangement of special devices, the essence of which I intend to set forth. I shall first note that the instantaneous closing of pipes is most often the result of the action of the turbine regulator, which, by virtue of the qualities required of it, must of necessity act very quickly. Some of the above-mentioned devices are directly connected with the regulator.

All devices in general may be divided into three categories:

1) devices that do not permit an instantaneous stoppage of the flow,
2) devices and arrangements that absorb the energy of the stopped mass,
3) devices that stop the supply of the pipe in the event of incipient destruction.

To the devices of the first category belongs the Escher Wyss arrangement, shown in drawing 8. This arrangement is shown applied to the Pelton wheel.

Cylinder A—the cylinder of the servomotor—under the action of the regulator controlling the outlet and inlet of water into the openings “a, a” of this cylinder, causes the piston p, moving up and down, to rotate part B, thereby increasing or decreasing the water discharge. When piston p moves, and consequently together with it part B moves downward, the right end of lever M rises upward, carrying with it cylinder C, filled with water, as a result of which opening m opens, differential piston k is moved upward by the difference of pressures, and the water finds an outlet along arrows F. Cylinder C, however, does not remain in place, but again seats itself downward, slowly passing water through the gap between the piston and the walls, whereby opening m is gradually closed; the pressure on piston k becomes greater in the downward direction, and it slowly descends, closing opening F.

The second device in drawing 9 belongs to the Bell firm. Here there are three hollow cylinders A, B, and C, inserted one into another; in the bottom of cylinders C and B ...

there are small hairlike openings. Under the action of the regulator controlling the water pressure in cylinder A, in the event of a reduction in work, both cylinders B and C move downward, covering the lip F; at the same time the outlet opening k, which discharges the water, opens, while spring M, fixed at one end in the immovable wall, tends to return piston C to its original position, moving it slowly upward until all the obstructing water has escaped through the hairlike opening. When the spring and piston C occupy their original position, opening k also closes. Both of the devices described are connected with the regulator.

The device shown in drawing 10 comes into action as a result of an increase of pressure in the pipe; here, under the action of the head, plate n bulges outward, opening m is uncovered, and the differential piston k comes into action, rising upward and discharging water along arrow F. With a decrease of pressure, plate n returns to its original position, opening m is covered, and piston k seats itself in place, slowly closing the outlet opening.

A spring valve made by the Rüsch works operates in a manner quite similar to the device just described. Its arrangement is shown in drawing 11. When the pressure rises, the spring is compressed and water is discharged; then, as the pressure decreases, the valve opening diminishes and finally closes completely.

To the second category of devices belongs the air vessel, which is placed at the end of the pipe; in this case the water, entering the vessel, compresses the air, and work is expended on this; but then the air tends to expand, and therefore prolonged oscillations of pressure occur in the pipe.

An improper calculation of the air vessel causes many difficulties; especially often there is observed a derangement of regulation as the result of unsuitable dimensions of the vessel. Since the theory of the air vessel appears insufficiently developed and confused, and since calculations are not always confirmed by practice, air vessels are now almost entirely not used by practitioners, at least in Switzerland, and are found for the most part in comparatively old installations.

Much more reliable than air vessels is the surge tower; this is an open tower of considerable diameter, which is installed at the end of the pipe; at the moment when the flow stops, the energy of the water goes into lifting the water in the tower, and then, as the velocity decreases, the level of the water in the tower descends; with such an arrangement the pressure in the pipe increases only by the height to which the water rises in the tower, which may be chosen arbitrarily, depending on the diameter of the tower. The tower, however, can be installed only on horizontal pipes or on pipes of small inclination.

To the second category may perhaps be assigned the device known under the name “brechplatte” (bursting plate), drawing 12.

This device consists of a short section of pipe with two flanges, the internal opening of which is not continued through, but a comparatively thin wall is left, so that if such a part is attached to the end of a water pipe, the end of the pipe is closed tightly; the plate must have such strength that, when the pressure rises by 10%–20% above normal, it breaks and the water finds a free outlet. By theoretical calculation, of course, it is impossible to guarantee exactly the limit of pressure that the plate will withstand; therefore these

the plates are made in the following way: first several pieces are made from a certain sort of cast iron, differing little in size from that calculated, and tests are made with them; then those which satisfy the test are selected as samples, and others are made according to them, and again tests are made at random; if, out of three taken, two withstand the test, then the third is considered good and may be used. Many such plates are made at once, since, after all, damage is rather frequent. The diameter of the plate must be calculated so that the quantity of water flowing out at the time of rupture should be no greater than that which would flow through the pipe before the sudden stop.

Since, at high pressure, the water flowing out of the pipes has an enormous velocity, dangerous for the integrity of objects it encounters, the pipe with the plate must be installed in such a way that the jet would not strike the walls of the premises. It is best to discharge the water along the canal, or, still better, into a great mass of water. After the plate has burst, the flow is stopped by a manual mechanism.

To the third category belong devices installed at the beginning of the penstock, such as the device shown in Fig. 13. If an expansion is made in the penstock and a heavy ball is placed in it, then, at low velocity and with the pipe in a horizontal position, the ball does not impede the motion; but if the velocity increases, then the ball will be carried along by the water and will close the pipe.

In this form, however, the device is applicable only for pipes of very small diameter.

Fig. 14 shows schematically a device whose action is the same as that of the preceding one; the weight g serves to balance the water pressure on valve A at normal velocity; when the velocity is increased the valve closes; in order to let the water pass again, one must press on the weight g, and thus set the valve back in its former place; since this device is installed on pipes of considerable size, in order to lessen the force required, when desired, to open the closed valve A, a bypass pipe with cock C is made for equalizing the pressure on both sides of the valve.

On the pipe, following the device just described, there must be installed a return air valve, which would admit air inward after the pipe has emptied; otherwise the external atmospheric pressure may crush the pipe.

On very large pipes, of diameter from 1 meter and more, this device, as practitioners complain, does not act especially properly. Escher Wyss builds safety devices of this same category set in motion by electricity. Let us have the inlet end of a pipe; (Fig. 15) on it is arranged a cover A, capable of closing shut; cover A is held suspended by the whole chain wound around a pulley; on the shaft of the pulley is fixed rigidly a fork C, held in place by a piece of soft iron placed in a coil; if current passes through the coil, then the iron is drawn inward, the fork C is released, the drum rotates, and cover A, under the action of its own weight and the pressure of the water, closes shut; the closing of the circuit is produced by a special wing installed in the pipe; it turns under the action of the increased velocity; the same closing of the circuit may also be produced, in case of danger, by the machinist located at the lower end of the pipe near the engines.

To illustrate the application of the devices I have described, I shall borrow a scheme from the project on the Albulа River. Drawing 16 represents this scheme in longitudinal section. The riverbed is dammed by a dam to a height of about 12 meters, forming a pond of 3,000,000 cubic meters; from this pond there emerges a water-conveying tunnel 7,240 meters long, which it is planned to drive in 8 sections, i.e. to begin work in 16 places simultaneously. The first 100 meters of the tunnel have an enormous cross-section, almost equal to that of a railway tunnel. In this part the velocity of the water does not exceed 0.3 meters; this makes it possible for sand and silt to settle to the bottom. Farther on follows a tunnel of 8 square meters cross-section, 7,140 meters long; the velocity of the current in it reaches 2 meters; at the outlet end of the tunnel a cylindrical reservoir is to be arranged, 15 meters in diameter and 12 meters high, counting from the highest water level in the dammed pond.

It is easy to calculate that, in the event of an instantaneous stoppage of the current, the work of the water lifted into the reservoir will be exactly equal to the living force of the water stopped in the tunnel.

For greater safety, an iron overflow funnel has been made in the reservoir, which, when the level rises, carries off part of the water. From the reservoir two iron pipes emerge, each 2 meters in diameter. At point M on these pipes manual sluice-gates and self-acting devices are installed, closing the pipe in the event of a dangerous increase of velocity in it, which may occur if the pipe bursts or ruptures somewhere lower down. At point N, where the turbine is installed, there are again manual sluice-gates and devices connected with the regulator, of the type assigned to the first category; in addition, on each pipe a “brechplatte” is installed.

Turbine Construction at the Factories

Before proceeding to a brief description of the factories I inspected, I shall first make the following remarks.

The emergence in Switzerland of large turbine-building factories took place with the appearance of the first hydroelectric stations, that is, comparatively recently. During this time turbine construction has made, and continues to make, very great progress, so great that theory has lagged far behind practice and is developing chiefly on the basis of types and installations already realized.

Every successful innovation increases one’s chances in the sense of competition, but for the most part only until theory has analyzed and worked it out; then, when the essence of the matter has become sufficiently clear, it is not difficult for another factory to construct something similar, even by circumventing the patent taken out by the first. Under such circumstances every leading factory tries carefully to guard its innovations; as a result, acquaintance with methods of calculation and access to these factories is extremely difficult for outsiders. Not long ago the calculation of Francis turbines of the high-speed type was a secret known only to a few factories; now, however, theory has disclosed this secret, and if anything in this subject still has to be concealed, it is only the technical methods of executing a problem solved by theory. Now, apparently, the turn has come for automatic turbine speed governors; their theory is still very new: for example, theoretical equations for governors of constant speed of the American type appeared

in the literature for the first time only last year, in the book by W. Bauersfeld, Die automatische Regulirungen der Turbinen. Berlin 1905.

In view of the exceedingly high level of the requirements at present imposed on electricians, and of the great complexity of the entire phenomenon in the regulator, long experience will probably never cease to play a prominent role in the construction of these devices, and its results constitute a secret carefully guarded at every plant.

The same may be said of many other subjects relating to turbine construction.

In view of all that has been said, access for a thorough inspection of a plant is very difficult; to be sure, if one appears directly at the plant, it is almost always possible to obtain permission merely to inspect it, i.e., in effect, to walk through all the departments of the plant together with the chief engineer. Such a tour lasts 1½–2 hours and, of course, can be of no benefit. Only a well-founded recommendation can somewhat mitigate this system, which seems to have been adopted at all the turbine-building plants in Switzerland; the recommendation must come from a person personally known to the plant. The person having such a recommendation may count on at least many of his questions being answered.

I succeeded in making a good inspection of the plant “Esher § Wyss” in Zurich, less fully—the plant “J. Ritter” in Winterthur and “Rüsch” in Dornbirn. To my written request for permission to inspect the plant “Bell” in Kriens I was refused; thus, apart from these three plants, I had to inspect only two small workshops, which can hardly be counted.

The plant “Echer § Wyss” enjoys a first-class reputation not only in Europe but also in America; installations of this plant are found in all parts of the world. The type of turbines currently being built by the plant consists exclusively of Francis turbines of every possible type and Pelton wheels of the American and European type. Girard and Jonval turbines are encountered only as taken in for repair or renewal. Two types of regulators deserve special attention: hydraulic—for large turbines, and hydromechanical—for turbines of medium size. Many installations have shown the high quality of these regulators; thus, with hydraulic regulators, at full load shedding it was possible to achieve a speed increase of only 3%–4%.

In producing the pattern for the blade die of a high-speed Francis turbine, the plant uses its own method, without resorting, as is generally accepted, to a series of parallel secant planes in order to reproduce the blade surface; instead, only 3–4 conical sections are taken. The high efficiency of the machines built at this plant is achieved not only by proper theoretical development, but also, in particular, by careful technical execution, often accompanied by a considerable increase in the cost of production. Thus, for example, nowhere except here did I see manual filing and scraping of the buckets of Pelton wheels, the passages of Francis turbines, and the passages of centrifugal pumps.

To perform such work, the turbine wheel is cast only with one rim a a (Fig. 17) and with the blades; the rim B B, however, is absent: it is cast separately and turned by means of a support moving along a template; in the same way

In the same way, the part of the edge of the blades that is to adjoin the ring \(B B\) is machined. The rim is fastened to the blades with screws.

I had occasion to see similar work being done on the turbine of Francis, cast in phosphor bronze; after removing the rim \(B B\), the workmen sawed the blades with a saw according to a template, and then brought them to the smoothness of glass with a scraper and emery paper.

Here I also examined thoroughly the molding of a wheel for a high-speed Francis turbine. On the horizontal bottom of the prepared pit (Fig. 18), the cone \(B\), representing the cavity of the suction tube, is set up; around the cone a plaster ring \(a\) is cast, on which the pitch of the blades is marked out. The iron blades, prepared beforehand, are placed on the marked ring and are fitted with one side to the cone \(B\). In order that they should stand firmly in place, after installation they are all tightened with an iron hoop or wire \(C\); then the entire space between the blades is filled with molding earth. The surfaces forming the boundaries of both rims \(m n\) and \(p q\) between the edges of the blades are prepared by an experienced molder by hand, with the aid of scrapers and smoothing irons; then the hoop \(C\) is removed and the previously prepared upper flask is put on. The edge of the blade that is to be poured into cast iron is serrated and driven into the core with a hammer; tinning of the edges is not practiced, and the edges are only carefully cleaned of rust on an emery wheel.

The Pelton wheel mold is changed by the works almost every year; in the pattern shop I counted more than 30 such typical molds. The steel and bronze buckets are also finished entirely by hand; the edges of these buckets are so sharp that I sharpened pencils on them several times.

In addition to all this, the works constantly tests its turbines and governors and introduces improvements and innovations every year. Besides hydraulic motors, the works also regards high-pressure centrifugal pumps, the “Zoell” steam turbine, and paper-making machines as its specialty; in addition, the works also undertakes all iron and boiler work.

The “Jacob Ritter” works in Winterthur is the first works to have begun building American-type turbines in Switzerland; it also has a worldwide reputation. The works is driven by turbines and has, moreover, an apparatus for brake testing of turbines, so that all innovations are immediately tested and evaluated. Among the outstanding features of the works one may note the special design of the regulating devices for Fink blades, which do not suffer from turbid water carrying much sand and silt; but the most outstanding specialty consists of turbine governors of all possible systems. Among them, the constant-speed governor is in general use, theoretically giving a constant turbine speed under different loads. In actual fact, however, at full unloading, changes in speed are obtained that do not exceed \(1\%\).

Very interesting, and probably of great importance for the purpose of saving water in our country in the Urals, is the constant-water-level governor. It is used for the most part where the turbine works in parallel with a steam engine, when, consequently, a speed governor is not installed on the turbine. The diagram of this governor is shown in Fig. 19. \(A\) and \(B\) are rows of continuously rotating, po-

receiving motion from the turbine shaft; shaft C, being joined by means of a worm to shaft B, also rotates continuously; on it slides cam m, held at a known height by the rocking lever f, to the other end of which is attached float K, floating on the water. In the position shown in the drawing, the cam, rotating, sets cylinders S, S in such a position that the fork rotating about axis X keeps the belt on the idle pulley; when the float rises or falls, eccentric m pushes back either the upper projection of one cylinder s or the lower projection of the other; each of the cylinders s has only one projection, the first of them at the top and the other at the bottom. Shifting the belt onto the right- or left-hand pulley from the idle one, shaft M receives rotation in one direction or the opposite, closing or opening the turbine channels.

The “Rüsch” works in Dornbirn is already in Austria, almost at the Swiss border; it does not have the worldwide reputation of the preceding two and, in technical terms, is significantly weaker. Nevertheless, it is supported by work chiefly for the local district. As a specialty of the works one may note a high-speed braking regulator, in which braking is achieved by increasing the operation of a centrifugal pump included as part of the regulator. According to reports, this regulator works very well in the sense of maintaining the turbine speed; unfortunately, like all braking devices, it is very uneconomical, and therefore it is convenient to use it only when there is an excess of water. In addition, the works has several of its own designs which, incidentally, represent nothing outstanding, but, on the other hand, this works operates more cheaply than the first two.

Plate

A.
Plan.
from the tunnel; grating.
Fig. 1.
Section along AB.
floor; threshold.

Plan.
upper; lower; A; B.

Fig. 4.
\(0.25\) to \(0.5\) m.

Fig. 3.
turbines; discharge sluice; threshold.

Fig. 5.
Plan with dimensions \(2.5\) m and \(2.5\) m.

Fig. 6.
Sectional drawing; visible dimensions \(25\) and \(35\) [[unclear: units not visible]].

R. Rhine.
dam; turbine canal; old channel.
Fig. 9.

Fig. 7.
red copper; \(c\); \(d\).

Fig. 8.
\(A\), \(B\), \(C\), \(K\), \(M\), \(m\), \(f\), \(f\).

Fig. 9.
\(A\), \(B\), \(G\), \(K\), \(M\), \(m\), \(f\), \(f\).

Fig. 10.

Fig. 11.

Fig. 12.

Fig. 13.

Fig. 14.

Fig. 15.

Fig. 16.

Fig. 17.

Fig. 18.

Fig. 19.