by Vaclav Smil
FIGURE 2.8. Drawing attached to Edison’s U.S. Patent 239,147 for a system of electric lighting (feeder and main arrangement). Reproduced from Edison (1881).
FIGURE 2.9. Edison’s Jumbo, a massive dynamo that was first displayed in Paris in 1881. Reproduced from Scientific American, December 10, 1881.
The project was originally intended as a European demonstration of the new system, and its location was chosen by Edison’s London representatives in order to circumvent the lengthy application process for digging up the streets by running electric lines under the Holborn Viaduct (Smithsonian Institution 2003). Machinery for this temporary installation was housed in a four-story row house (No. 57), with a 93-kW generator supplying initially 938 lamps, including 164 street lamps along 800 m of the viaduct and lights in the General Post Office, the City Temple, and businesses along the street, with direct current (DC) of 110 V. A second dynamo was added by April 1881, and the station remained in operation until 1884. But Edison’s plans for a large central plant in London were thwarted by the Electric Lighting Act of the British Parliament, which made such installations impossible until it was revised in 1888.
Just as the Manhattan station was undergoing a series of final tests, Edison’s first American small hydroelectric station, powered by a 107-cm waterwheel, was readied for service on the Fox River in Appleton, Wisconsin. Two small dynamos rated at a total of 25 kW were housed in a wooden shed, and they powered 280 weak lights (10 candle power), but the sturdy installation, ordered by the town’s paper manufacturer H. F. Rogers, was in full operation by September 30, 1882, and it worked until 1899 (Dyer and Martin 1929). The first small English hydro-generating plant, opened during the previous year in Godalming by the Siemens brothers, whose company installed a small water turbine in the River Wey and laid cables in the gutters, operated only until 1884 (Electricity Council 1973).
FIGURE 2.10. Dynamo room of Edison’s New York Pearl Street Station with six directly driven Jumbos. Reproduced from Scientific American, August 26, 1882.
The Manhattan station was to be in an entirely different class. Edison looked for a dilapidated building to site his generating station in a cheap property and ended up with two houses, 255 and 257 Pearl Street, both much more expensive than he anticipated. The station’s heavy equipment was put in No. 257, coal (and ash) in the basement, four Babcock & Wilcox boilers (about 180 kW each) on the ground floor, and six Porter-Allen engines (each rated at 94 kW) and six large direct-connected Jumbo dynamos on the reinforced second floor (Martin 1922; figure 2.10). The neighboring house, No. 255, was used for material storage and repair shop and, frequently, as a dormitory for Edison and his crew working on the station. Edison devotion to the project included not only a close supervision of different tasks but actual work in the trenches (Israel 1998).
Even before all connections were completed, Edison turned on the first light in J. P. Morgan’s office at 3 P.M. on September 4, 1882: just a single dynamo was online, and it supplied about 400 lights. When the time came to engage the second machine, the usually confident inventor admitted that he was extremely nervous as he scheduled a Sunday test—and, as he recalled later, his concerns were justified:
One engine would stop and the other would run up to a thousand revolutions; and then they would seesaw. The trouble was with the governors… I grabbed the throttle of one engine and E. H. Johnson, who was the only one present to keep his wits, caught hold of the other, and we shut them off. (Edison cited in Martin 1922:56)
To another witness,
it was a terrifying experience, as I didn’t know what was going to happen. The engines and dynamos made a horrible racket, from loud and deep groans to a hideous shriek, and the place seemed to be filled with sparks and flames of all colors. It was as if the gates of the infernal regions had been suddenly opened. (Cited in Martin 1922:56-57)
By the end of 1882, three more Jumbos were added at the Pearl Street station, whose output was lighting more than 5,000 lamps, and in January 1883 the company began charging for its electricity using Edison’s first ingenious but cumbersome electrolytic meters. Edison had every reason to be pleased. As he noted in 1904 in an article he wrote for Electrical World and Engineer, “As I now look back, I sometimes wonder at how much was done in so short a time” (cited in Jehl 1937:310). The scope of his early work during those years is perhaps best revealed by the patents he obtained between 1880 and 1882 in addition to nearly 90 patents on incandescent filaments and lamps: 60 patents dealing with “magneto or dynamo-electric machine” and its regulation, 14 patents for the system of electric lighting, a dozen patents concerning the distribution of electricity, and 10 patents for electric meters and motors (TAEP 2002).
Higher than anticipated costs and frequent outages of dynamos meant that the company lost money in 1883 ($4,457.50), but it made a good profit ($35,554.79) in 1884 (Martin 1922). Its operation provided a unique learning opportunity and served as an irreplaceable advertisement for the new system. One of the first adjustments Edison made soon after starting the Pearl Street station, on November 27, 1882, was to file a patent for the three-wire distribution system (U.S. Patent 274,290), an arrangement that was independently devised in England by John Hopkinson (1849-1898) a few months earlier and that remains the standard in our electric circuits.
Edison’s first central station with the three-wire distribution was completed in October 1883 in Brockton, Massachusetts. This configuration saved about two-thirds of the copper mass compared to the two-wire conduit used in Manhattan (figure 2.11). While the feeder-and-main system is still with us, the three-wire DC transmission was not destined to be one of Edison’s enduring innovations. In a few years it was superseded, both in its maximum spatial reach and its unit cost, by the transmission of three-phase and single-phase AC, which was initially shunned by Edison but strongly favored by Westing-house, Tesla, Ferranti, and Steinmetz.
By 1884 the Pearl Street station was serving more than 10,000 lamps, and its success led to a rapid diffusion of similar installations for which the Edison Company had a virtual monopoly during most of the 1880s. By 1891 more than 1,300 central Edison plants were in operation in the United States, supplying about 3 million lights. Fire that damaged the Pearl Street station in 1890 destroyed all but one of the Jumbos, No. 9; in 1893 it was moved to the Columbian Exposition in Chicago, and it was eventually rebuilt and ended up in the Ford Museum in Greenfield Village near Dearborn, Michigan (Sinnott and Bowditch 1980; ASME 2003). The gutted station was reopened in just 11 days (Edison joined the repair crews in the round-the-clock effort to restart the operation), but it closed down in 1895, and the building was later demolished.
FIGURE 2.11. Two of seven drawings illustrating Edison’s specification of his U.S. Patent 274,290 for three-wire electrical distribution. The left image shows wiring attached to a dynamo; the right, to secondary batteries. Reproduced from Edison (1883).
Edison’s role in the genesis of the electric era cannot be overestimated. He was able to identify key technical challenges, resolve them by tenacious interdisciplinary research and development, and translate the resulting innovations into commercial use. There were other contemporary inventors of lightbulbs and dynamos, but only Edison had the vision of a complete system as well as the determination and organizational talent to make the entire system work, and he and his coworkers translated his bold ideas into realities in an astonishingly short period. One of the greatest tributes paid to this work came from Emil Rathenau, one of the creators of Germany’s electric industry and the founder of Allgemeine Elektrizitats Gesselschaft, on the occasion of his 70th birthday on December 11, 1908.
Rathenau (quoted in Dyer and Martin 1929:318-319) recalled his impressions after seeing Edison’s display at the Paris Electrical Exhibition of 1881:
The Edison system of lighting was as beautifully conceived down to the very details, and as thoroughly worked out as if it had been tested for decades in various towns. Neither sockets, switches, fuses, lamp-holders, nor any of
the other accessories necessary to complete the installation were wanting; and the generating of the current, the regulation, the wiring with distribution boxes, house connections, meters, etc., all showed signs of astonishing skill and incomparable genius.
This praise alone negates all the disparaging retelling of history and many derisory remarks that Edison’s antagonists and critics have been producing since the very beginning of the electric era. Edison was, without any doubt and to resort to more of the modern parlance, an extraordinary systems thinker, but it seems to me that Friedel and Israel (1986:227) captured best the essence of his achievements by noting that “the completeness of that system was more the product of opportunities afforded by technical accomplishments and financial resources than the outcome of a purposeful systems approach.” This more subtle interpretation does not change the basic facts. Edison was an exceptionally inventive, ambitious, and confident man who awed, motivated, inspired, and alienated his coworkers and financial backers. The combination of unusual insights, irrepressible enthusiasm, and the flair for (self)promotion led him often to voice exaggerated expectations and to make impossible promises.
Perhaps the most notable example of these attitudes was reflected in headlines a New York Sun reporter (cited in Friedel and Israel 1986:13) chose in September 1878 after talking to Edison at what was the very beginning of the search for practical incandescent light: “Edison’s Newest Marvel. Sending Cheap Light, Heat, and Power by Electricity. ‘I have it now.’” All that at a time when he actually had none of these capabilities! But obstacles and setbacks that derailed his overoptimistic schedules served only to fortify his determination to overcome them and to come up with better solutions. His devotion to the pursuit of invention was legendary, his physical stamina incredible. What is not perhaps stressed enough is that the financial support he received from his backers (including some of the richest men of his time) and the skills and talents of his craftsmen made it possible for him to explore fairly freely so many ideas and possibilities. In that sense, the Menlo Park laboratory was a precursor of the great corporate R&D institutions of the 20th century.
In retrospect, there is no doubt that the combination of Edison’s inventions or radical improvements of several key components of the emerging electric system and his indefatigable push for its commercialization have been his greatest legacies. His were epoch-making, truly monumental, achievements, but I must hasten to note that while all of our electric networks are conceptual descendants of Edison’s system of centralized electricity supply, their technical particulars as well as their operational arrangements had changed substantially before the end of the 19th century. Unlike the remarkably durable specifications and external features of Edison’s incandescent lightbulbs, Edison’s original electric system was not such a resilient survivor—and it did not deserve to be.
Above all, as advanced as it was for a pioneering design, Edison’s first station was still a very inefficient generator of electricity: its heat rate was about 146 MJ/kWh, which means that it was converting less than 2.5% of the coal it burned into electricity. By 1900, the two key components of the electricity-generating system were different: steam turbines connected directly to alternators, rather than steam engines coupled with dynamos, became the preferred prime movers, and increasingly higher voltages of AC, rather than relatively low voltages of DC, distributed the generated electricity. On the consumption side, improved incandescent lights still used a large share of the generated electricity, but electric motors in factories and in urban transportation were rapidly becoming the largest consumers.
Turbines, Transformers, Motors
Incredibly, during the remainder of the 1880s the new electric industry saw advances that were no less fundamental than those made during the first three years of its development when it was dominated by Edison’s race to introduce a profitable commercial system. These changes took place at every stage of the innovation process as they transformed the generation of electricity, its transmission, and its final uses. Edison remained an important, but increasingly a marginal, player, and the greatest acclaim must deservedly go to two engineers born in the mid-1850s: Charles Algernon Parsons (1854-1931) for his invention of steam turbogenerator, and Nikola Tesla (1856-1943) for his development of polyphase electric motor. Although William Stanley’s (1858-1916) contributions to the design of transformers stand out, genesis of that device was much more a matter of gradual refinements that were introduced by nearly a dozen engineers in several countries.
Every one of these three great innovations made a fundamental difference for the future of large-scale electricity generation, and their combination amounted to a fundamental transformation in an industry that was still taking its very first commercial steps. By 1900, these advances came to define the performance of modern electric enterprises, determining possible size and efficiency of individual turbogenerating units, location and layout of stations, the choice of switchgear, and transmission and distribution arrangements. Without these innovations, it would have been impossible to reach the magnitude of generation and hence the economies of scale that made electricity one of the greatest bargains during the course of the 20th century.
All of these techniques were later transformed by series of incremental innovations whose pace had notably accelerated during the post-WWII economic expansion but then reached performance plateaus during the early 1970s. Consequently, all components and all processes of modern electric industry are now more efficient and more reliable than they were in 1914, yet they demand less material per unit of installed capacity and are more economical, and their operation has reduced impacts on the environment. At the same time, there is no mistaking their 1880s pedigree.
When Edison began to outline his bold plan for centralized electricity generation for large cities, he had only one kind of prime mover to consider: a steam engine. During the penultimate decade of the 19th century, design of these machines was a mature art (Thurston 1878; Ewing 1911; Dickinson 1939), but they were hardly the best prime movers for the intended large-scale electricity generation. After more than a century of development, their best efficiencies rose to about 25% by 1880, but typical sustained performances of large machines were much lower, closer to 15%. And although the mass/power ratio for the best designs fell from more than 500 g/W in 1800 to about 150 g/W by the 1880s, steam engines remained very heavy prime movers.
In addition to relatively low efficiency, high weight, and restricted capacity, top speeds of steam engines were inherently limited due to their reciprocating motion. Common piston speeds were below 100 m/min, and even the best marine engines of the 1880s, although they could reach maxima of up to 300 m/min, worked normally at 150-180 m/min (Ewing 1911). Moreover, this reciprocating motion had to be transformed into smooth rotation to drive dynamos. Edison’s first generators installed in the experimental station in Menlo Park in 1880 were driven, most awkwardly and least reliably, by belts attached to steam engines. This arrangement was soon substituted by direct-connected high-speed engines, and Edison used the best available models in his first installations in 1881. A Porter-Allen machine he ordered was capable of 600 rpm but had problems with its speed regulator. An Armington & Sims engine for the Paris exhibition was rated at about 50 kW and had 350 rpm (Friedel and Israel 1986).
Simple calculations reveal the limitations inherent in these maxima. Imagine an urban system that does not serve any electric motors but only 1 million lights (merely six lights per household for 100,000 families and another 400,000 public, office, and factory lamps) whose average power is just 60 W. Such a system would require (assuming 10% distribution losses) 66 MW of electricity. With the best dynamos converting about 75% of mechanical energy into electricity, this would call for steam engines rated at 88 MW. Those at Edison’s Pearl Street station had the total capacity of nearly 3.4 MW, which means that our (dim and frugal) lighting system for a city of 400,000 people would have required 25 such generating stations. Their load factor would have been l
ow (no more than 35%) as they had only one kind of final use, but even with relatively high efficiency of 20% they would have consumed annually about 250,000 t of steam coal.
This was, indeed, the inevitable setup of the earliest urban electric systems: they served just limited parts of large cities, and hence the generating companies had to bring large quantities of coal into densely populated areas and burn them there. Switching from DC to AC would have allowed for efficient transmission from peri-urban locations, but it would not have changed the required number of generating units. Moreover, this setup was so inefficient that were the entire U.S. population (about 57 million people in 1885) to be served by such low-intensity, systems they would have consumed the country’s total annual production of bituminous coal during the mid-1880s! There was an obvious need for a more powerful, but a much less massive and more efficient, prime mover to turn the dynamos.
Steam Turbines: Laval, Parsons, Curtis
Water-driven turbines have been able to do this since the 1830s when Benoit Fourneyron introduced his effective designs, but before the 1880s there were not even any serious attempts to develop an analogical machine driven by steam. The first successful design was introduced by a Swedish engineer Carl Gustaf Patrick de Laval (1845-1913), whose most notable previous invention was a centrifugal cream separator that he patented in 1878 (Smith 1954). His impulse steam turbine, first introduced in 1882, extracted steam’s kinetic energy by releasing it from divergent (trumpet-shaped) nozzles on a rotor with appropriately angled blades. Such a machine is subjected to high rotational speeds and huge centrifugal forces. Given the materials of the 1880s, it was possible to build only machines of limited capacity and to run them at no more than two-thirds of the speed needed for their best efficiency. Even then, a helical gearing was needed to reduce the excessive speed of rotation.