PART TWO: POWER AND ENGINEERING 372 the safety of the system he arranged a demonstration in which a concentric cable of his own design was made live at 10,000 volts and a workman drove a chisel through it. The workman was unhurt and the Board of Trade were satisfied. The main conductors were concentric copper tubes separated by paper impregnated with ozokerite wax. They were entirely successful, some remaining in service until 1933. The generators, however, were less successful and, to make matters worse for Ferranti, the Board of Trade would only allow the company to supply a much smaller area of London than he had hoped. The directors decided to build only a small part of Ferranti’s planned generating plant, and in 1891 he resigned from the Corporation. The Deptford scheme was very ambitious, and although the subsequent history of electricity supply shows that Ferranti was working on the right lines, he may well have been more immediately successful if his plans had been rather more modest. All public electricity supply systems are now entirely AC, although the last DC supply in Britain—to a Fleet Street newspaper—remained in use until 1985. For transmission over long distances DC systems are sometimes preferred. Very high power mercury arc valves, developed in Sweden in the 1930s, permit very heavy currents to be rectified, transmitted over a DC line, and then inverted back to AC. DC transmission is also used when it is desired to link two AC systems which operate at different frequencies or which operate at the same frequency but cannot be kept in synchronism. An example of the former is in Japan, where part of the country operates on 50Hz and part on 60Hz, and the two networks are interconnected through a DC link. The cables under the English Channel linking the British and French grids also operate on DC. The first Anglo-French link, opened in 1962, had a capacity of 160MW. New cables, laid in 1986, with a capacity of 2000MW, are buried in a trench cut in the sea floor to avoid damage by ships’ anchors or fishermen’s trawls. ECONOMICS: OFF-PEAK LOADS At first the public electricity supply was used almost exclusively for lighting, but this meant that the system was not used to its maximum efficiency because the demand was concentrated in to a few hours each day. In 1888, Crompton produced a graph of the demand for electricity over an average day in which he showed that the load exceeded 50 per cent of the peak load for only five hours out of the twenty-four, and exceeded 75 per cent of peak for only about two hours. The peak load determined the plant costs, and the plant had to be manned all day and all night. Any load that could be attracted at times other than the peak hours could therefore be supplied economically at a price barely greater than the fuel cost. ELECTRICITY 373 The potential sources for such off-peak loads were electric heating and cooking, and industrial uses. Electric cooking was vigorously promoted, by advertising and by articles in the press. Most supply undertakings offered electricity at reduced price—often half-price—for cooking and hired out appliances at low rentals. A school of electric cookery was opened in Kensington, London, in 1894 to promote the idea. In 1912, A.F.Berry opened his ‘Tricity House’ which combined a restaurant serving 700 meals a day with a showroom for cooking equipment. Berry also gave lectures on the advantages of electric cooking, in which he argued that one ton of coal in a power station could do as much cooking as ten tons of coal delivered to the house (see Chapter 19). The reduced prices for electricity used other than for lighting continued into the 1930s. Today off-peak electricity is still available in the UK at about half- price, but lighting is now only a fraction of the total demand and the off-peak hours are in the night. MEASUREMENT The rapid progress of electric lighting in the 1880s, created a demand for practical and reliable measuring instruments. Most important were devices for measuring voltage and consumption. Accurate voltage measurements were essential because the life of filament lamps was critically dependent on the supply voltage. Devices to measure electricity consumption were required so that customers could be charged according to the electricity they had used. The first scientific attempts at electrical measurements were in the eighteenth century. Electrostatic forces between charged bodies were measured as early as 1786 when the Revd Abraham Bennet wrote a letter to the Royal Society describing the gold leaf electroscope. In this two pieces of gold leaf hanging side by side are in electrical contact with a conductor whose potential is to be observed. Since the gold leaves become similarly charged they repel each other, and hang apart at an angle which is a measure of the potential. The most important of the early workers on electrostatic measurements was Charles Augustin de Coulomb, who invented the torsion balance electrometer and used it to establish the inverse square law relating the force of attraction between charged bodies and the distance between them. Coulomb’s instrument had a long silk thread supporting a horizontal straw inside a large glass cylinder. The straw was covered with sealing wax and the silk thread was fixed to a cap at the top of the cylinder which could be turned. Also inside the cylinder was a metal ball connected to an external knob which could be charged. The device was so sensitive that Coulomb could measure a force of a few milligrammes. PART TWO: POWER AND ENGINEERING 374 More immediately relevant to the practical business of electricity supply, in the late nineteenth century Sir William Thompson designed electrostatic voltmeters which relied on measuring the force of attraction between charged elements of the instrument. Two patterns were in widespread use. The quadrant electrometer, which was sold commercially from 1887, had two vertical sheets of metal with a narrow space between them into which a pivoted vane was drawn by electrostatic attraction. This instrument could measure up to 10,000 volts. His multi-cellular meter, made from 1890, had what were essentially a number of ‘quadrant’ movements side by side on the same axis, and these meters were used for voltages of a few hundred. A modified form of the quadrant electrometer, with additional vanes, was used as a wattmeter. Until the relatively recent advent of electronic measuring instruments, most current measurements have depended on measuring the magnetic force created by a current in a wire. This magnetic force has been balanced against a controlling force, which may be produced by a spring, by gravity, or by another magnet. Simple galvanometers, in which a large diameter coil surrounded a pivoted magnetic needle, with the earth’s magnetic field providing the controlling force, were made in 1837 by Claude-Servain Pouillet. Such instruments were accurate and straightforward to use, but they were bulky and the needle took a long time to come to rest, so measurements were slow. The electricity supply industry wanted instruments that were portable, quick to read, and stable in the readings they gave. In the 1880s there were no really stable magnetic materials, and simple instruments making use of a magnetized needle had to be recalibrated from time to time. A current measuring instrument which did not depend on the vagaries of magnetic materials was the electrodynamometer. This measured the forces between current-carrying conductors, so it was really only suitable for high current work. Being very reliable, however, it was frequently used in laboratories and for calibrating other instruments. The principle of the electrodynamometer is due to Wilhelm Weber, who was following up Ampère’s mathematical study of the forces between currents. A moveable coil, usually supported by a torsion suspension, hangs within a fixed coil. When current flows in both coils magnetic forces tend to twist the moveable coil. Usually the moveable coil is returned to its original position by twisting the top of the suspension through an angle which is a measure of the current. An instrument of this kind was used to monitor the operation of the generator in the first public electricity supply system, at Godalming. A laboratory version of the electrodynamometer was the Kelvin Ampere Balance, in which the attractive force between fixed and moving coils was balanced by a sliding weight on a balance arm. The arrangement was similar to a balance for weighing. In 1894 two such instruments were held by the Board of ELECTRICITY 375 Trade as legal standards for the measurement of current. The accuracy was stated to be within 0.2 per cent—a matter for great satisfaction at the time! The need for calibration, and all the problems attendant on changing calibration, can be avoided if the instrument is used simply to determine the presence or absence of a current. This is the principle of the potentiometer and bridge measuring circuits. In the potentiometer, which was first described in 1841, a constant current was passed through a straight uniform wire placed alongside a linear scale. The potential between two points on the wire was therefore proportional to the distance between them, and this could be read from the scale. Once the scale had been calibrated by means of a voltage standard, other voltages could be measured accurately. The unknown voltage was connected through a galvanometer to a length of the potentiometer wire, and when the galvanometer showed that no current flowed then the unknown voltage was equal to the voltage across that length of wire. The potentiometer was used in the bridge circuit described by Wheatstone in a long Royal Society review paper on electrical measurements in 1843. (Although it has been known ever since as the Wheatstone bridge, he clearly attributed it to S.H.Christie.) The bridge circuit was intended for comparing resistances. An unknown resistance and a standard were connected in series, carrying the same current, and in parallel with a potentiometer. A galvanometer was used to determine the point on the potentiometer wire at which the voltage was the same as at the junction of the two resistances, and the ratio of the resistances was then the same as the ratio of the two sections of the potentiometer. The rapid progress in electric lighting around 1880 created a demand for direct reading instruments. The new filament lamps were far more critically dependent on the supply voltage than the older arc lamps, which could tolerate wide fluctuations without damage. The first practical moving iron ammeter was devised by Professors W.E.Ayrton and John Perry in 1879. It had an iron core which was drawn into a circular coil carrying the current to be measured. In the ‘magnifying’ version the core was restrained by a spring wound helically from flat metal strip. As the spring was stretched the free end rotated, turning the pointer. Other moving iron instruments of the period included Schuckert’s, in which an asymmetrically mounted piece of soft iron was made to twist by the magnetic field within a current-carrying coil. The moving coil meter, with a light coil hanging in the annular air-gap between the poles of a permanent magnet and an iron cylinder, was devised by J.A.d’Arsonval. Such instruments were first made on a large scale by the Weston Electrical Instrument Company of New Jersey. Weston introduced the idea of ‘damping’ the movement by mounting the coil on a copper former, so that its motion was damped by eddy currents and it came to rest quickly when a reading was taken. He also used jewelled PART TWO: POWER AND ENGINEERING 376 bearings to reduce friction. The unipivot instrument, in which the moving parts are carried on a single bearing to give further reduced friction and greater sensitivity, was introduced by the London instrument-maker R.J.Paul in 1903. Hot-wire instruments make use of the fact that a wire expands when heated. Philip Cardew designed the first commercial hot-wire instrument in 1883. It was reliable and accurate, but rather unwieldy, having three metres of platinum-silver wire stretched over pulleys in a tube almost a metre long. One end of the wire was fixed, the other held taut by a spring and connected by gearing to a pointer. The hot-wire principle was adapted by Hartmann and Braun in the 1890s to produce a more manageable instrument. They utilized the sag of a taut wire rather than its change in length. The pointer of their instrument was driven by a wire attached to the middle of the current-carrying wire and pulling it sideways. Arthur Wright, the engineer of the Brighton electricity undertaking (see p. 370), made the first recording meter about 1886 to monitor the load on his system. He used a strip of paper coated with lampblack which was pulled along by clockwork and marked by a pointer on a simple moving iron ammeter. Commercial electricity supply required meters to measure the energy consumed, so that customers could be charged accordingly, although a few early systems simply charged according to the number of lamps connected. Edison made electrolytic meters, in which the current to be measured passed through a cell containing two zinc plates in zinc sulphate solution. The plates were weighed periodically to measure the transfer of zinc and hence the current that had flowed. The first direct reading supply meter was devised by Ayrton and Perry, though usually known as the Aron meter after the man who improved and then manufactured it. These instruments made use of the fact that the speed of a pendulum clock depends on the force of gravity. Two pendulum clocks were coupled together through a differential gearing connected to dials which showed the difference in ‘time’ measured by each clock. Coils were arranged adjacent to the iron pendulum bobs so that the apparent force of gravity varied in accordance with the supply being measured. Most energy meters have what is in effect an electric motor driving a disc which is restrained by an eddy current brake. The motor torque is arranged to be proportional to the electrical energy passing; the braking torque is proportional to speed. The total number of revolutions of the disc is then a measure of the energy consumed. Elihu Thomson, born in England but brought up in the USA, devised the first motor meter in 1882. He founded the Thomson-Houston company in 1879, jointly with his former teacher E.J.Houston, to make arc lighting equipment, but they soon expanded into the whole range of electrical manufactures. Ferranti’s mercury motor meter ELECTRICITY 377 had a copper disc rotating in a bath of mercury, with current flowing radially in the disc. The Edison, Thomson and Ferranti meters described were essentially DC instruments (although Thomson’s would work on AC). After the invention of the induction motor (see p. 384) induction type instruments were adopted for all AC systems and are now used in virtually all electricity supplies. With alternating current systems there was interest in the nature of the wave-forms involved. A simple contact operated synchronously with the supply by a cam made it possible to monitor the supply voltage (or current) at any specific point in the cycle, and by taking a series of such measurements it was a simple though tedious process to build up a picture of the waveform. Wheatstone had developed the method in the course of his telegraph researches, and the idea was reinvented by Joubert in 1880 when studying the behaviour of arc lighting circuits. A mechanical oscillograph which would give a visual display of a complete waveform was suggested in 1892 but only achieved in 1897 when William du Bois Duddell succeeded in making a galvanometer whose movement was light enough to follow the variations of an alternating current waveform. In another field, medicine, such instruments were developed to study the electrical action of the heart. The discovery of the electron in 1897 led to the cathode ray tube. Tubes were made which produced a beam of electrons on a screen and could deflect the beam in two axes at right angles. Circuits were developed to give a deflection varying with time on one axis. The voltage being studied was then applied to deflect the beam in the other axis, and a complete waveform drawn out on the screen. A review of such instruments published by the Institution of Electrical Engineers in 1928 gave more space to mechanical than to cathode ray oscilloscopes, but during the 1930s the cathode ray instrument completely replaced the mechanical. The first internationally agreed electrical standards were drawn up at the International Congress of Electricians that met in Paris in 1881. Before that many workers used their own standards, though the British Association had been considering national standards for some years. The Congress defined units for resistance, current and voltage, and since that time electrical science and engineering has benefited from universally accepted standards of measurement. This work is now the responsibility of the International Electrotechnical Commission, in Geneva. ELECTROMAGNETIC ENGINES The story of the electric motor really begins with Oersted’s discovery in 1819 that a compass needle could be deflected by an electric current. In 1821, Michael Faraday showed that it was possible to produce continuous rotary motion by PART TWO: POWER AND ENGINEERING 378 electromagnetism (see p. 355). During the nineteenth century many machines were designed that produced mechanical motion from electromagnetic effects, collectively known as ‘electromagnetic engines’. One early machine was made by William Sturgeon in 1832, and it was probably the first electromagnetic engine to be put to practical use—to turn a roasting spit. Sturgeon’s machine had a vertical shaft carrying two compound permanent magnets, one at the top, one at the bottom, with their north poles pointing in opposite directions. As the shaft turned, the ends of the permanent magnets passed close to the poles of the four vertical electromagnets fixed on the base board. The commutator was an elaborate arrangement with two concentric mercury cups carried round by the shaft, into which wires dipped, and a horizontal disc cut into four quadrants, with wiper arms pressing on them. Another early but little known maker of electromagnetic engines was Sibrandus Stratingh, a medical doctor and Professor of Chemistry at Groningen in the Netherlands, who wanted to make an electric road vehicle. In 1835 he constructed a table-sized model, but he never achieved a full-size version. Like a number of other early inventors, however, he did make an electric boat and he managed to take his family in a boat, electrically powered, in 1840. The first person to obtain patents for electromagnetic engines was Thomas Davenport. He patented a machine in the USA in 1837, and later the same year obtained an English patent also. A model of his machine, now in the Smithsonian Institution in Washington, has a rotor consisting of four coils on a cruciform frame fixed to a vertical shaft. Opposite pairs of coils are connected in series and the ends of the wires go to simple brushes which press on a two- part commutator consisting of two semi-circular pieces of copper. The battery is connected to the copper pieces. The stator is two semi-circular permanent magnets with their like poles adjacent. Another American inventor was W.H.Taylor, who exhibited a motor in London in 1840. Taylor’s machine was written up enthusiastically in the Mechanics Magazine (see Figure 6.9). The construction was quite simple. An arrangement of four electromagnets on a frame surrounded a wooden wheel with seven soft iron armatures around its edge. A simple commutator on the axis switched on each of the four electromagnets in turn. Taylor claimed that earlier ideas for electromagnetic engines had depended on reversing the polarity of electromagnets. He said that his invention was the idea of switching the magnets so that they were simply magnetized and demagnetized, but not reversed in polarity. It seems that he realized that it took a significant time to reverse the polarity of an iron-cored electromagnet. The Scotsman Robert Davidson made motors which also operated by switching the electromagnets on and off, not reversing them. In the winter of 1841–42 the Royal Scottish Society of Arts gave him financial help with his experiments, and in September 1842 he made an electrically driven carriage which ran on the Edinburgh and Glasgow railway. The four-wheeled carriage ELECTRICITY 379 Figure 6.9: Engraving of Taylor’s electromagnetic engine. Reproduced from Mechanic’s Magazine, 9 May 1840. Figure 6.10: Electromagnetic engine by Wheatstone at the time Taylor’s machine was exhibited. PART TWO: POWER AND ENGINEERING 380 was nearly 5m long and weighed about 5 tonnes. The motors were wooden cylinders on each axle with iron strips fixed on the surface. There were horseshoe magnets on either side of the cylinder which were energized alternately through a simple commutator arrangement on the axis. The batteries, a total of 40 cells each with a zinc plate between two iron plates, were ranged at each end of the carriage. The plates could be raised out of the wooden troughs containing the electrolyte by a simple windlass. The carriage could run at about 6.5kph (4mph) on level track. There is no contemporary record of the distance actually travelled by the vehicle: presumably it did not actually travel all the way from Edinburgh to Glasgow. In 1839 the Tsar gave a grant to Professor M.H.Jacobi of St Petersburg for work on an electric motor, probably the first government grant ever given for electrical engineering research. In a letter to Michael Faraday, Jacobi described how he had arranged for an electromagnetic engine to drive the paddlewheels on a boat and travelled for days with ten or twelve people aboard on the River Neva. When supplied from a battery of 128 Grove cells, the vessel travelled at about 4kph. In some cases enough data is given in contemporary records for the efficiency of these machines to be calculated, and figures of 10 to 20 per cent are obtained. The efficiency of electromagnetic engines was possibly a matter of some interest in the 1840s. In 1843, Charles Wheatstone was describing a variety of electrical devices in a paper to the Royal Society. One was the rheostat, which developed initially as a measuring device, but Wheatstone said that it could be used for controlling the speed of a motor; he also said, wrongly, that a rheostat in series with a motor could control its speed without any loss of efficiency. In 1849 the United States Commissioner of Patents, Thomas Ewbank, included some thoughts on the subject of electric motors in his annual report to Congress. He said: ‘The belief is a growing one that electricity…is ordained to effect the mightiest of revolutions in human affairs.’ He referred to various experiments with electric motors and then continued, somewhat pessimistically, ‘but these experiments, interesting as they certainly were, have brought no marked results, nor afforded any high degree of encouragement to proceed. It might be imprudent to assert that electromagnetism can never supersede steam, still, in the present state of electrical science the desideratum is rather to be hoped for than expected.’ Ewbank’s pessimism was not shared by Congress. In 1850 the United States Congress gave $20,000 to Professor Charles Page of Massachusetts, to develop electromagnetic engines, apparently with the navy mainly in mind. In a report to the Secretary of the Navy, Page said that he had made machines of one and four horsepower and asked for a grant to build a machine of 100hp. The most ingenious of all the early electromagnetic engines must surely be Allan’s machine made in 1852. This is basically a reciprocating engine with four cranks and four ‘piston rods’. Each piston rod carries four armatures ELECTRICITY 381 which press on collars on the rod but are not otherwise fixed to it. There are 16 sets of coils, one for each armature, and the coils are energized one at a time by a commutator. Each electromagnet is therefore energized for 1/16 of a revolution. As each armature reaches its electromagnet it is stopped by it, while the piston rod continues its travel. All these machines were, in any commercial sense, failures. They were entirely dependent on expensive chemical batteries, but if that were the only reason then the electromagnetic engine would have flourished when generators made electricity readily available in the 1880s. In modern terminology these engines were all ‘magnetic’ machines which depend on direct magnetic attraction between stator and rotor. Most modern motors, such as the induction motor (see p. 384), are electromagnetic machines in which the fields on one side of the air gap generate currents on the other. Conventional machine theory tells us that magnetic machines get better as they get smaller, while electromagby about 1880, although no theoretical reasoning had been set out. PRACTICAL ELECTRIC MOTORS The fact that the machines used as generators could be used also as motors was recognized quite early. The Siemens brothers, for example, noted in 1872 that the ‘small rotating machine runs just as well as a motor as it does as a generator’. The firm of Siemens & Halske set out to find customers requiring electric power transmission. The first public exhibition of the electrical transmission of power by means of a generator and motor was probably the demonstration at the Vienna Exhibition in 1873. There the Gramme Company showed two identical machines linked by wires but 500m apart; one was used as a generator, the other was used as a motor driving a pump. Similar demonstrations were given at other exhibitions in the USA and in Britain in the following few years. By 1874, Gramme had electrically-driven machinery in his Paris factory—though he used a single large motor driving a line shaft, as in a steam powered factory, not an individual motor for each machine. Several exhibitors at the first International Electrical Exhibition, held at Paris in 1881, demonstrated motors. Marcel Deprez, for example, had a motordriven sewing machine, lathes, a drill and a printing press. Siemens exhibited a lift within the building and a tram running towards it along the Champs-Elysées. One of the first Gramme machines to be sold as a motor was used to drive a conveyor for sugar beet in a factory at Sermaize, France. It was so successful that in May 1879 the factory owners, Messrs Chrétien & Félix, decided to try electric ploughing. The plough was hauled across a rectangular field by ropes between two wagons, each carrying a motor and a winding drum. . resistances, and the ratio of the resistances was then the same as the ratio of the two sections of the potentiometer. The rapid progress in electric lighting around 1880 created a demand for direct. to be observed. Since the gold leaves become similarly charged they repel each other, and hang apart at an angle which is a measure of the potential. The most important of the early workers on. which operate at the same frequency but cannot be kept in synchronism. An example of the former is in Japan, where part of the country operates on 50Hz and part on 60Hz, and the two networks