by Vaclav Smil
Lamps with tantalum filaments (melting point of 2,996°C) made from a drawn wire reaching efficacies up to 7 lm/W were patented by Werner von Bolton and Otto Feuerlein in 1901 and 1902 and used until 1911. In 1904 Alexander Just and Franz Hanaman patented the production of tungsten filaments; the metal’s high melting point (3,410°C) allowed efficacy of at least 8 lm/W, but preparing ductile (bendable) wire from this grayish white (hence its German name, Wolfram, and symbol W) lustrous metal was not easy (MTS 2002). Several methods were tried during the following years before William David Coolidge (1873–1975) demonstrated that the brittleness is due to the grain structure of the element. By using high temperature and mechanical treatment, tungsten becomes so ductile that it can be drawn into a fine wire even at ambient temperature; his patent application for making tungsten filaments was filed on June 19, 1912 (figure 2.6).
General Electric’s first tungsten lamp, with the filament in vacuum, was introduced in 1910, and it produced no less than 10 lm/W (efficiency of about 1.5%). The metal became the dominant incandescent material by 1911, and the last cellulose lamps were made in 1918. When incandescing inside lamps, both filaments must be kept at temperatures far below their respective melting points (carbon just above 2,500°C and tungsten below 2,700°C) in order to avoid rapid vaporization. Consequently, their energy output peaks in the near infrared (at about 1μm) and then falls through the visible spectrum to about 350 nm (into the ultraviolet A range), producing most of the visible light in red and yellow wavelengths, unlike daylight, which peaks at 550 nm and whose intensity declines in both infrared and ultraviolet directions.
FIGURE 2.6. Two of six illustrations that accompanied Coolidge’s U.S. Patent 1,082,933 for making tungsten filaments show his wire-drawing apparatus and an incandescent lamp made with the ductile metal.
One more step was needed to complete the evolution of standard incandescent light: to eliminate virtually all evaporation of the filament and thus to prevent any black film deposits inside a lamp. That step was taken in April 1913 by Irving Langmuir (1881-1957), who discovered that instead of maintaining a perfect vacuum, it is more effective to place tungsten filaments into a mixture of nitrogen and argon and, in order to reduce the heat loss due to convection, to coil them (U.S. Patent 1,180,159; see figure 1.8). These measures raised the efficacy to 12 lm/W for common lightbulbs and more for the lamps of high power. Less than a quarter-century after Edison’s 1879 experiments, the production of incandescent lamps was thus a mature technique. To be sure, gradual gains during the 20th century—improvements such as coiled tungsten coils (introduced in 1934) that further reduced heat loss from convection—brought further gains. Efficacy rose to more than 15 lm/W for 100-W lamps, and rated life spans increased to 1,000 hours for standard sizes, but no fundamental changes were ahead (figure 2.5).
By the 1990s more efficacious lights, based on entirely different principles, had captured large shares of the illumination markets. Although most of these lights also trace their origins to the pre-WWI period, they became commercially successful only after 1950. The first discharge lamps were built by Peter Cooper Hewitt (mercury vapor in 1900), and Georges Claude (1870-1960) demonstrated neon discharge (bright orange red) in 1910 (first neon lights were used in 1912 at a West End cinema in London). Further development of these ideas led to commercial discharge lamps: low-pressure sodium lamps were introduced in 1932, and low-pressure mercury vapor lamps, commonly known as fluorescent lights, were first patented in 1927 and came on the market during the late 1930s (Bowers 1998).
In fluorescent lamps, phosphorous compounds that coat the inside of their glass absorb ultraviolet rays generated by excitation of low-pressure mercury vapor and reradiate them as illumination approximating daylight. The best fluorescent lights are now producing nearly 110 lm/W, converting about 15% of electricity into visible radiation—and they also last about 25 times longer than does a tungsten filament. In 1912 Charles Steinmetz experimented with metal halide compounds in mercury lamps to correct their blue-green color—and exactly 50 years later General Electric revealed its first commercial metal halide light (whose successors now produce up to 110 lm/W).
Although the earliest lightbulbs were only as luminous as gas jets that they were designed to replace (typically equivalent to just 16 candles or about 200 lm), their light was safer, more reliable, quiet, and less expensive, which is why they rapidly displaced even Welsbach’s greatly improved gas mantle. By 1914, 100-W tungsten lightbulbs were able to produce more than 1,000 lm, and during the coming decades they became the principal means of bringing daylight to billions of people. Remarkably, even the wide choice of non-incandescent lights that have been commercialized since the 1930s and that offer much higher efficacies and much longer life spans has not turned lightbulbs into rare relics of a simpler era.
Despite the inroads made by more efficient sources, incandescent lightbulbs continued to dominate the lighting market throughout the 20th century. The most detailed study of the U.S. residential energy consumption showed that during the early 1990s, 87% of all lights used one or more hours per day (453 million out of a total of 523 million) were incandescent (EIA 1996). And hundreds of millions of poor villagers in Asia, Africa, and Latin America whose dwellings have no connections to electric networks still wait for that soft light to change their lives as much as it has transformed the habits and opportunities of the world’s more affluent minority. Those smallish fragile glass containers with an incandescing filament may not be around by the end of the 21st century—but they were the first source of convenient and flexible light that gave us, finally, an easy mastery over darkness, undoubtedly one the most common, most recognizable, and most beneficial inventions of the pre-WWI period that helped to create the 20th century and keep it brightly lit.
Edison’s System
Invention of commercially viable lightbulbs had obviously a great practical and symbolic importance in ushering the electric era, but the bulbs were, both physically and figuratively, just the end of the line. Edison’s lasting contribution is not that he set to invent a lightbulb (in that quest he was, as we have seen, preceded by a score of other inventors) or that he succeeded, relatively rapidly, in that effort. The fundamental importance of Edison’s multifaceted and, even for him, frenzied activity that took place between 1879 and 1882 is that he put in place the world’s first commercial system of electricity generation, transmission, and conversion. In Hughes’s (1983:18) words,
Edison was a holistic conceptualizer and determined solver of the problems associated with the growth of systems…Edison’s concepts grew out of his need to find organizing principles that were powerful enough to integrate and give purposeful direction to diverse factors and components.
Few complex technical systems have seen such a rapid transformation of ideas into a working commercial enterprise. But the first step was taken reluctantly. One of the businessmen present at the Menlo Park demonstration on December 31, 1879, was Henry Villard, the chairman of Oregon Railway & Navigation Co., who became an instant convert and persuaded reluctant Edison, who preferred to concentrate on the development of a larger scale urban system, to install electric lights in his latest steamship. And so the first Edisonian system was put onboard the Columbia beginning in March 1880. The ship left for Portland in May 1880 and arrived in Oregon, after a 20-week journey around the South America, with its lighting in excellent condition. But this success was not followed immediately by any larger land-based installations.
This is understandable given the amount of technical, economic, and managerial challenges that had to be overcome. Electricity had to be cheaper than gas lighting, a mature, well-established industry with decades of experience, with extensive generation and transmission infrastructures in place in every major Western city and with profitable operations owned by some of the leading investors of that time. By the early 1880s, customers were used to the gently hissing sound of burning gas, but the experience was made less comfortable by the evolved heat and e
missions of water vapor and carbonic acid. While electric light of 100 candles generated just 1.2 MJ of heat an hour and released no emissions, Argand gas burners of equivalent luminosity warmed the surroundings at the rate of about 20 MJ/hour and emitted nearly 1 kg of water vapor and nearly half a cubic meter of carbonic acid (Anonymous 1883; Paton 1890). Moreover, there were dangers of asphyxiation or explosion from leaking gas, and, obviously, the burner could be used only in an upright position.
Greater convenience of electric lights was obvious: steady, nonflickering, noiseless, nonpolluting, and flexible, ready to be used in all positions. But, as illustrated by Siemens’s opinion cited in chapter 1, not everybody agreed that the days of gas lighting were numbered. And regardless of one’s beliefs about the competitiveness of the two systems, it was very difficult to project the costs of electric system as there was no commercial production of most of the key components needed for its functioning and, naturally, no operational experience that would reveal the frequency of breakdowns and the overall system reliability to deliver electricity to customers. Planning far ahead, months before he had the first long-lasting lightbulb, Edison set aside his work on incandescent filaments and plunged not only into designs and construction of larger dynamos but also into detailed studies of operating costs, profits, and pricing of the coal gas industry. These studies were needed to set the goals that his system had to meet in order to prevail.
Dynamos, Engines, Fixtures
After his return from a memorable trip to California in August 1878, Edison began testing two of the most successful dynamos of the day, machines by Werner Siemens (1816-1892) and Zénobe-Théophile Gramme (1826-1901) that were used to power arc lights. By that time dynamo design advanced far beyond the first, hand-cranked, toylike generator built by Hypolite Pixii in 1831 (MacLaren 1943). By far, the most critical gain came in 1866-1867 with the realization that residual magnetism in electromagnets makes it possible for the generators to work even from a dead start, that is, without batteries and permanent magnets. In his memoirs, Werner Siemens (1893; figure 2.7) recalled his first presentation of the idea of what he named “dynamo-electric machine” to a group of Berlin physicists in December 1866 and its publication on January 17, 1867, as well as the fact that the priority of his invention was immediately questioned but confirmed later. But there is enough evidence to conclude that Charles Wheatstone (1802–1875), Henry Wilde, and Samuel Varley discovered the same phenomenon independently during the same time.
FIGURE 2.7. Werner Siemens, a founder of one of the world’s leading makers of electric (and now also electronic) equipment, inventor of self-exciting dynamo and builder of first long-distance telegraph lines. Portrait reproduced from Figuier (1888).
In an 1883 lecture, Siemens’s brother William (1823-1883) noted that “the essential features involved in the dynamo-machine…were published by their authors for the pure scientific interest attached to them without being made subject matter of letters patent,” and that this situation “retarded the introduction of this class of electrical machine” because nobody showed sufficient interest in the requisite commercial development (Siemens 1882:67). That is why Siemens gave a great credit to Gramme for his initiative in using Antonio Pacinotti’s 1861 idea to build a ring armature wound with many individual coils of wire insulated with bitumen and also to introduce a new type of commutator.
Gramme’s invention of a new machine magneto-electrique produisant de courant continu was presented to the Academie des Sciences in Paris in July 1871 (Chauvois 1967). In contrast to Gramme’s hollow cylinder, Siemens’s improved dynamo had windings crossing near the center. Both of these machines could supply continuous current without overheating, were used to energize arc lights in an increasing number of European cities of the 1870s, and were later replaced by more powerful dynamos designed by Charles Brush and Elihu Thomson. Edison began his systematic work on producing better dynamos in February 1879, in parallel with the much better known work on incandescent lights, by examining closely the existing designs. He eventually adopted the drum armature, but the laboratory spent a great deal of time on devising new wiring patterns and designing new commutators and magnets.
The most distinguishing feature of dynamos that were actually used in the earliest lighting demonstrations was the use of two oversized, polelike, field magnets that earned them the nickname of “long-legged Mary-Ann” (Friedel and Israel 1986). Remarkably, these machines had efficiency as high as 82%, but a much larger dynamo was needed for the contemplated urban system capable of powering more than a thousand lights. Its design began in the summer of 1880, and its most distinct feature was the direct connection to a Porter-Allen steam engine, obviating the inefficient belting. The prototype could eventually energize 600 lamps of 18 candlepower, proving the basic viability of the concept.
The first installation of underground distribution network on the grounds of Menlo Park laboratory in spring and summer of 1880 uncovered problems with the existing insulation materials. The chosen wrapping consisted of layers of muslin with paraffin, linseed oil, and tar, and the first street lights supplied by underground conduits were lit on November 1, 1880. Edison’s house was connected to the network a week later. Optimistic as ever, Edison now predicted he will have a working central station in the lower Manhattan before May 1, 1881. He needed to move fast, because by December 1880 no fewer than six companies were installing arc lights in the city, and because Hiram Maxim put in the first incandescent lights in the vaults and reading rooms of the Mercantile Safe Deposit Co. (Bowers 1998). But first, a new Edison Electric Illuminating Co. of New York, set up by Edison’s attorney Grosvenor Lowrey, had to be formed (in December 1880 with the initial capital of $1 million) to conform to a state law that restricted the use of street lights to enterprises incorporated under gas statutes.
Then a deal had to be made between Edison and the company, controlled by the investors, for sharing the eventual profit, and all of the system’s components had to be assembled. Armington & Sims supplied the steam engines, and Babcock & Wilcox provided aid on the boilers, but everything else— switch boxes, fittings, sockets, wall switches, safety fuses, fuse boxes, consumption meters, and insulated underground conductors—had to be designed, tested, improved, and redesigned. As Frank Lewis Dyer (the former general counsel for Edison’s laboratory) recalled, the leading manufacturer of gas-lighting fixtures, Mitchel, Vance & Co.,
had no faith in electric lighting and rejected all our overtures to induce them to take up the new business of making electric-light fixtures… Mr. Edison invited the cooperation of his leading stockholders. They lacked confidence or did not care to increase their investment. He was forced to go on alone. (Dyer and Martin 1929:718-719)
Once some of these new factories began turning out fairly large numbers of electric components, Edison abandoned his resistance to install small, isolated systems in individual plants or offices. In February 1881 the first installation of this kind was completed in the basement of Hinds, Ketcham & Co., a New York lithography shop. Many similar installations followed during the rest of 1881, and in November of that year the Edison Company for Isolated Lighting was organized to handle new orders coming from universities, hotels, steamships, and textile mills, where the electric lights were particularly welcome in order to reduce the risk of fire. By the beginning of 1883, there were more than 150 isolated systems in the United States and Canada and more than 100 in Europe. But Edison’s main goal was a large centralized system in Manhattan, whose realization was falling behind the previous year’s expectation. Finally, the contract for this project was signed on March 23, 1881.
The First Central Plants
Edison initially wanted to supply the entire district from Canal Street on the north to Wall Street on the south but had to scale the coverage to about 2.5 km2 between Wall, Nassau, Spruce, and Ferry streets, Peck Slip, and the East River. The location, the First District in lower Manhattan, was selected for a combination of reasons, above all its high density
of lighting needs and its proximity to New York’s financial and publishing establishments. Edison’s crews surveyed the area’s lighting needs, as well as its requirements for mechanical power, already during the year 1880 and then produced large maps annotated in color inks that showed Edison the exact number of gas jets in every building, their hours of operation, and their cost (Dyer and Martin 1929).
The laying of more than 24 km of underground conductors began soon after the contract was signed, and it proceeded according to Edison’s feeder-and-main system for which he was granted a patent (U.S. Patent 239,147) in March 1881 (figure 2.8). This arrangement cost less than 20% than would have the tree system of circuits, reducing the cost of copper from the originally calculated $23.24 per lamp to just $3.72. Meanwhile, the work continued on a larger dynamo intended first for powering Edison’s exhibits at the forthcoming Paris International Electrical Exposition (Beauchamp 1997).
Charles Clarke, Upton’s college classmate and the first Chief Engineer of the New York Edison Co., had a major role in the development of a new design whose success was assured once the resistance of its armature was reduced to less than 0.01 П and the problems with air cooling were ironed out (figure 2.9). The massive machine (nearly 30 t) of unprecedented power (51.5 kW at 103 V) achieved public notoriety because of its nickname, Jumbo, derived from the fact that it was shipped to France on the same vessel that on a previous voyage transported P. T. Barnum’s eponymous elephant from the London Zoo (Beauchamp 1997). By the end of 1881, the second Jumbo was installed at Edison’s first operating central station in London, which began generating electricity on January 12, 1882.