An AC Pioneer
United Electric Light & Power Company
It often appears that everything that could be written has been written with regard to the 1890s competition between alternating current (ac) and direct current (dc) electric power systems. Much is inaccurate and, unfortunately, is repeated ad infinitum. One factor that all writers, both popular and scholarly, overlook is that the issue was not resolved in a single event nor was it immediate. The ac systems that ultimately prevailed required decades of research and development prior to universal application.
The Early Days
Leading experts, including legendary mathematician and electrical engineer Charles Steinmetz, believed ac distribution to be impractical in urban areas of dense load concentration. Their reasons included the lack of battery reserve, poor voltage regulation, the inefficiency of ac motors, and the limited understanding of reactive power management and related ac phenomena.
Others believed those obstacles could be resolved. The United Electric Light and Power Company of New York (United) was destined to be a foremost leader in the adoption of ac for urban loads. Founded in 1887 as the Safety Electric Light and Power Company (Safety Electric), the company was approached by George Westinghouse, the nation’s most dedicated proponent of ac development. His Westinghouse Electric and Manufacturing Company of East Pittsburgh, Pennsylvania, held patents critical to the commercial development of ac, the transformer patent of William Stanley, and also the polyphase system and induction motor patents of Nikola Tesla. Westinghouse also had rights to significant European patents.
Safety Electric management, uncertain of a system considered impractical by many, expressed skepticism but licensed a Westinghouse system. Reorganized under Westinghouse control, it took the United name on 9 December 1889 and was merged with the United States Illuminating Company (USI), an arc light firm with significant franchises and corporate connections. Over time, the most important USI contribution was Frank W. Smith, a manager who had started as a 12-year-old office boy in 1880. United initially installed a Westinghouse 1,000-V, 133-Hz, single-phase lighting system in lower Manhattan. Each customer was supplied through a transformer connected to overhead lines. Despite the inefficiency of many small distribution transformers, the company persevered.
Following the forced removal of overhead lines, by 1895 United had leased a total of 135 mi (217.3 km) of underground ducts throughout Manhattan in a bold display of confidence in the future of ac distribution. Customers were limited to the area below 59th Street, but major technical change came with adoption of the Westinghouse two-phase system to supply 60-Hz power for motors as well as lights. The power was generated by a pair of 1,500-kW alternators installed in a reconstructed arc light plant on 29th Street at the East River. These machines had been rebuilt after removal from the 1893 Columbian Exposition in Chicago, the site of the first large-scale Westinghouse installation.
As United sought to expand the market, the East 29th Street property was enlarged into a full-scale power plant that measured 197 ft (60 m) × 160 ft (48.8 m), one of the largest of that era. At that time, the majority of United customers were located along the waterfront areas as the Edison Electric Illuminating Company had focused on the central portion of the island. Arc lamp illumination of streets remained a major business as well, and United acquired the Brush Electric Company in 1892. By 1895, United’s revenue totaled US$500,000 annually.
New, direct-connected alternators (claimed to be the first without belt drive) went into operation at 29th Street in 1896 as lines were extended northward. On 26 December 1899, United lines reached the northern end of Manhattan to supply a customer on Dyckman Street. Company leadership aggressively pursued new approaches to the extent of the exploration of a steam turbine as a prime mover for alternators. A Westinghouse-Parsons unit was ordered but never installed, possibly because of impending corporate change. It was installed instead at a plant in Hartford, Connecticut. Electric franchises in and around New York City had been acquired by utility magnate Anthony N. Brady to achieve the economy of large-scale operation, a reorganization that established a new role for United.
United Established as a Leader
As United came under the control of the Consolidated Gas Company in 1900, Brady’s plan placed operational and technical decisions under the gifted engineer and inventor Thomas E. Murray. Smith continued to rise through the ranks of United management and became Murray’s associate in development plans. Smith shared Westinghouse’s belief in the ultimate superiority of ac, and he believed that United would be the keystone of 60-Hz ac systems in New York City. Unlike other Manhattan franchises, it was not merged into the Edison Company although it shared property and facilities.
In 1906, Electrical World magazine remarked upon the success of United in a variety of areas. A record industrial ac load included 174 motors in the National Biscuit Company plant on the lower west side. Midtown customers often used huge lighted signs, and the largest was the animated Heatherbloom Petticoat “Silk’s Only Rival” sign with 1,837 lamps that depicted a girl in a storm, her skirts battered by wind and rain. Mounted upon a 55-ft (16.8-m) wide frame were the outlines of a 70-ft (21.4-m) tall skirt with a 50-ft (15.2-m) wide bloom; the shoes measured 12 ft (3.7 m) long with 3-ft (0.91-m) buckles.
United sought residential customers with newspaper ads backed by billboards and banner signs at major street intersections. Small merchants were provided with free electric signs. The signs were expensive, with multiple sockets and ornate metal scroll work detail, but the cost was recovered through increased power consumption. Market surveys had determined that those merchants would not otherwise have installed electric signs. A free planning kit and promotional booklet titled “Lights and Shadows” was distributed to customers, architects, contractors, and realtors.
As forecast load surpassed the capacity of the 29th Street power station, additional power was made available through frequency and phase-changing motor-alternator sets that received 25-Hz, three-phase power from the New York Edison Waterside station. The city took title to the 29th Street site for the construction of Bellevue Hospital in 1907, which made United dependent on Waterside power. A pair of 16-MW, 60-Hz, two-phase alternators were installed in Waterside as a dedicated supply. The frequency and phase-changing motor alternator sets were relocated to New York Edison substations and connected to available United feeders.
A new United power station had been scheduled two years prior, but various issues delayed construction. In late 1913, United opened a modern power station at 201st Street on an inlet of the Harlem River named Sherman Creek. That location was selected in recognition of the northward shift of the United system load center, a move depicted in an Electrical World comparison of statistics from 1909 and 1913, as shown in Table 1.
Sherman Creek supplied power not only to United customers and those of affiliated Consolidated Gas companies as far as 30 mi (48.3 km) away, but also through contract to the New York, New Haven, and Hartford Railroad for which dedicated 25-Hz alternators were installed. System supervision remained at the Edison Waterside station, apparently transferred there upon closure of the 29th Street plant.
The first 60-Hz, two-phase system of 1895 connected 2,300-V feeders from 29th Street to primary distribution points from which lines radiated to the individual transformers that supplied single-phase “lighting” power and two-phase power for motors. When Sherman Creek opened, new 7,500-V (described as 7,800 V at the plant bus) three-phase lines connected to substations in which T connected Scott transformers provided the 2,300-V, two-phase power to the primary distribution points from which the lines radiated. Thus, new transmission lines and substations were overlaid on an existing system.
The substations were located along the length of Manhattan, one in a company garage and warehouse on West 146th Street and two others in former arc light plants on West 24th Street and Elizabeth Street. New compact structures designed to fit city lots were erected on 187th, West 97th, and West 45th Streets (see Figure 1). The Elizabeth Street property was subsequently rebuilt to the new standard. All substations were attended as distribution regulation was their prime function, that being obtained by feeder switches to balance transformer loads. Synchronous condensers were used when needed to supply reactive power and stabilize power factor.
West 187th Street was the largest substation as it supplied all of northern Manhattan. West 45th Street carried a late evening peak demand from theaters and advertising signs. Elizabeth Street showed a late afternoon industrial peak, while the others experienced the early evening peak common to residential neighborhoods. In most cases, multiple routes were used for as many as four feeder sets to protect against outages.
As United continued to expand, the distribution system became more complex. United engineers worked closely with those of Westinghouse to evolve a simpler system since what was in place could not be extended indefinitely. By 1920 there were a total of 70 area distribution points, with each supplied over a dedicated feeder from a substation and interconnected by “manhole” switches to enable a fault to be bypassed. A total of 250 such interconnecting switches were in place. Although this was adequate for the United system, which then supplied about 10% of the utility power distributed on Manhattan island, the scheme was by no means a replacement for the New York Edison dc distribution system.
United continued to add customers while new products were promoted in United Service, a monthly magazine with color covers and inserts sent to United customers (see Figure 2). Selected appliances were offered at a discount. Illustrated articles detailed dozens of new industrial, commercial, and residential customers from the lower west and east sides and modern luxury apartments on the upper west side and in northern Manhattan. For some, ac power possessed a definite advantage. One was the Erbograph Company on West 146th Street, a manufacturer of movie film. The process at that time carried a high risk of explosion, so the sparkless ac induction motor was an attractive option that United promoted vigorously in its industrial advertising.
During the 1920s, United conducted a number of advertising campaigns in New York City’s daily morning and evening newspapers. Each campaign included a series of advertisements promoting a particular theme. Figures 3—5 show representative advertisements published during three such advertising campaigns.
The sign business continued to grow as new multicolor animated signs appeared. The Wrigley’s gum sign was considered the world’s greatest with 15,000 lamps and changing colors. It was a full block in length, stood 34 ft (10.4 m) high, and spelled out three product lines alternately: “Spearmint,” “Doublemint,” and “Juicy Fruit.” Public monuments supplied by United included Grant’s Tomb in Riverside Park. United also provided back-up lighting power for the tunnels of the Interborough subway with circuits isolated from the transit company’s power system.
Displays at the annual electric show hammered the theme of the efficiency of ac equipment with “United Alternating Current Service” spelled out in lights. Billboards announced that modern ac appliances “Save Work; Preserve Youth” with bright illustrations of new appliances wielded joyfully by a young lady. Electric clocks developed by the Warren Self Winding Clock Company were promoted with a 10-ft (3.4-m) diameter outdoor clock on the 146th Street office of United. The clock kept perfect time, regulated by the alternators at the power station. Those in turn were regulated by a master clock set by telegraph signal from the U.S. Naval Observatory.
Superpower was a buzzword of the day, describing systems that supplied a large region rather than just a city. A second station, Hell Gate, located in the Bronx opened in November 1921 and brought United full status as a superpower system as it supplied power to affiliated companies in the Bronx, Queens, and Westchester (see Figure 6). Small plants of low efficiency in those areas were retired. For a time, Hell Gate held the status of the world’s most powerful steam-operated power station.
United sales for April 1922 tell a remarkable story. The incomplete Hell Gate station, built at a cost of US$18 million, supplied 40 million kWh with a peak load of 103.5 MW. The older Sherman Creek station, which cost US$6 million in pre-WWI dollars, supplied 33.5 million kWh with a peak load of 94 MW for a total of almost 200 MW peak load with total sales of 73.5 million kWh. That was a stupendous achievement for a relatively small company confined to an island with only 82,150 customers, 56,277 of which were primarily residential in northern Manhattan. Moreover, it distributed only 60-Hz ac in a totally urban environment, a unique achievement as the concept still sought full acceptance in the utility industry.
By 1922, United distributed 15% of the metered utility power consumed on Manhattan Island, and the balance was dc from New York Edison Company substations in its franchise area south of 135th Street. Utilities across the nation had pursued improved ac distribution as the investment in dc substations and the cables required by the heavy currents of 120/240-V dc distribution soared. A variety of ac distribution schemes were developed but all had limitations that prevented standardization. United operated only in urban territory and thus took the lead in the perfection of ac distribution.
Distribution Revolution: The United Automatic Network
The goal was a reliable system with reduced installation and operational costs that did not “saturate” neighborhoods with ducts and manholes. Though radial feeders worked well with loads up to 300 kVA, nonuniform loads prevented the operation of distribution transformers at the design load, which increased reactive power and reduced power factor. In the event of an outage, the hand-operated manhole switches produced lengthy delays of up to 30 minutes or more as crews had to reach the manhole before the fault could be isolated and bypassed. Reliability could be improved only through the use of duplicate feeders, a costly and inefficient method at best.
The interconnection of distribution lines in a network was deemed the most practical resolution and for that a set of what were called “essentials” was determined. A network had to be of practical (limited) size, adequate for peak load and reasonable future growth, with transformers spaced reasonably, and with voltage regulation equal to that of dc distribution. If substations were to be retained, the feeder voltage to the area transformers would have to be as high as practical.
After those requirements were met, what were termed “desirable” goals were set: lower first cost, with the flexibility to supplant any dc system with greater reliability. (The allegedly perfected dc systems in major cities had started to fail to such an extent that failure had been the subject of a national conference.) An ac network would have to maximize simplicity in terms of numbers of ducts, feeders, and junction box components. The long-term goal was to supplant the ac substation with an automatic network at a lower cost than the cumbersome dc distribution system.
An ac network would be three phase as new theorems (Fortescue and Clarke) made the balancing of single-phase loads on three-phase systems practical. The ideal operating voltage had to be determined. United had reportedly used, in some locations, a three-wire, two-phase system that supplied 110 V for lighting and 191 V for motors, later 115 V and 199 V, respectively. Westinghouse engineers reviewed the use of 115/199 V, 120/208 V, or 125/216 V. It was determined that lighting was best above 115 V while motors operated best below 220 V, as the slight increase in load current reduced the relative percentage of reactive power to improve power factor and reduce heat. The present standard of 120/208 V, three-phase was thus selected.
To assure continuity (reliability) equal to dc required that a network permit rapid restoration by allowing low-voltage secondary faults to “burn clear” without a shutdown and, if necessary, isolation of a high-voltage feeder without compromise to the rest of the system. Automatic reclosure of protective relays had to be limited to prevent “pumping” (repeated cycling open and closed) when a fault was not cleared.
In 1922 United developed the “New York (or United) automatic network” after ten years of effort. It mandated a single set of mains for both power and light with voltage regulation such that the starting of motors did not impact lighting. It was fully compatible with components on the market and was compatible with the high primary transmission voltage of the power stations to allow future elimination of ac substations. It interlaced primaries and was designed to “burn off” secondary faults without compromise to the system. Single conductor cables ensured that most faults would be to ground and thus “burn off” with less damage than phase-to-phase faults. Intended to maximize continuity, automatic network protection prevented “pumping.”
The ideal component location was a building vault that did not require expensive submersible transformers. If that was not practical, a sidewalk manhole location was preferred to the street to improve access and reduce the impact of traffic. Transformers would be located according to anticipated load centers unless the location was at an intersection where multiple access paths would be a priority. The latter anticipated the future connection of networks to the high-voltage primary power station lines without substations.
“Moment of Truth”
The first test installation was located to support a residential load adjacent to the West 97th Street substation. Rated at 300 kW, four transformer locations supplied apartment buildings containing 15 elevators. Voltage regulation allowed for a maximum variation of ±3% and no visible change of incandescent lights upon the start of motors. Load balance could be adjusted by isolation of a transformer when load was reduced. Transformer impedance was set at 8.7% to assure an equal load on each.
The moment of truth came on 12 April 1922 when the new system was cut over in place of a radial system. Any attempt to accomplish so much held the possibility of unanticipated problems, but the engineers of United and Westinghouse had planned well. Precise logs of all events through April 1923 revealed no major difficulties, with most incidents being attributable to component problems. The only major change was an increase in transformer impedance to 10%. Both past and recent histories document that network as the first fully successful ac automatic secondary distribution network, the one that set the standard for all that followed.
The network was installed rapidly throughout the system. Substation feeders to distribution transformers were increased to 3,000 V prior to 1922, most likely around the time the transmission voltage of 7,500 V was increased to 13,200 V (13,800 V in some reports). Networks had then to accommodate both substation feeders (2,300 and 3,000 V) and also 13-kV power station transmission lines. A subsequent development was the “spot” network for large institutional users and tall structures. As network installation expanded, it slowed and then reversed expansion of the complex radial feeder system of United, which had reached 2,000 mi (3,219 km) of cable by 1922.
The goal of the direct connection of networks to power stations without substations was accomplished in 1926. As networks were installed, three phase soon predominated with 65 MW of United’s 83 MW of distribution converted by 1928, a year pivotal in the long history of the ac revolution.
Power companies across the United States had begun replacement of dc distribution by the early 1920s, some with ac radial feeders and others with various ac network concepts. Electrical World projected, in 1923, that only Boston, New York, Chicago, and Philadelphia would retain dc in the downtown areas. The latter two cities began plans for a “changeover” (as the process was termed) later that year. New York Edison continued to extend the dc system until it reached a total of 41 substations with 282 rotary converters in 1928. Plans were underway for additional substations even as long-time dc proponent John Lieb acknowledged, at the dedication of the New York Edison East River (25 Hz) station in 1926, that a change to 60-Hz ac was on the horizon.
Matthew S. Sloan, upon becoming president of New York Edison, sought standardization. After an extensive review of the dc distribution, commenced in 1927, the results were announced on 20 November 1928. With the words “In light of the reliability, economy, and efficiency of the automatic A.C. network perfected by the United Company, there is no justification for continued extension of D.C. service,” Sloan issued a series of orders to retire dc service at company expense.
The competition was ended; the controversy silenced; the issue settled. Thirty-five years after the celebrated events at the Chicago Exposition of 1893 and the subsequent installation at Niagara Falls, ac had become the dominant system foreseen by Westinghouse, Smith, and the others. That victory came about not with great machines or stupendous structures but through the ingenious application of relays hidden in underground vaults and manholes.
Ironically, the very name United Electric Light and Power Company, the pioneer that achieved the success some believed to be impossible, has been long forgotten. The changeover at New York Edison was long, and the last dc utility entrance and meter were retired on 14 November 2007. Sloan had projected a period of 35–45 years to complete the changeover of all customers as the estimate of installed investment in dc customers’ property was US$100 million in 1928 dollars. To assist the changeover, United and New York Edison merged their technical departments in 1932. Three years later United was merged into New York Edison by transfer of stock, and the 46-year-old company ceased to exist. The merged corporation, New York Edison Company, Inc. merged with parent Consolidated Gas Company in 1936 to become Consolidated Edison Company of New York, Inc. (Con Edison). Later authors referred to the innovation as “the New York Edison network,” apparently unaware of the role of United. More recent technical papers, however, give credit where it is due, to that small company that defied conventional wisdom to turn a goal into a reality.
This article came about after comments and questions that followed publication of my previous article; “Architect of Power: Thomas E. Murray & New York’s Electrical System,” which cited United extensively. I realized that United itself is another piece of lost history. Having lived in prime United territory during my high school and college years, I had never heard of the company although I do recall a United logo on a meter panel. If I thought about it at all, I probably assumed it referred to a contractor or component supplier.
While doing research in the 1990s, I again encountered the name but knew little of it. Old articles and company reports revealed the long-forgotten primacy of the role of United in power distribution. Articles in Electrical World lavished praise upon United, while a masterpiece article by H. Richter in the August 1925 issue of Electric Journal helped unravel the Gordian knot that was the mystery of network evolution. Questions remain however. There are references to looped-through feeders but where is uncertain. References also mention 4,160-V feeders but details are few.
United supplied the Columbia-Presbyterian Medical Center in 1928 by connection to both the 146th Street and the 187th Street substations as a protection against complete outage. It would appear that 146th Street was supplied from Hell Gate, while 187th Street remained on Sherman Creek lines, presumably with transmission lines between both plants. Again details are minimal. In 1938, flood waters shut down the boilers at Hell Gate to black out major portions of the Bronx and Westchester plus former New York Edison dc territory in Manhattan. This flooding did not interrupt power to northern Manhattan and older sections of former United territory, which appear to have been supplied primarily by Sherman Creek. In later years after the merged systems were integrated fully, Hell Gate supplied former Edison dc territory directly. On 17 August 1959, a Hell Gate feeder failure shut down Manhattan from the south side of 110th Street to 59th on the west but only to 86th Street on the east.
While much has been lost or forgotten in the years since United ceased to operate under its own name, I was rewarded a few years ago with a casual comment from a long-retired Con Edison employee. In recounting his first days on the job at Sherman Creek around 1946, he referred to the old timers then who had been there at the beginning as “the United guys.” Someone actually knew! Hopefully this story will shed a little light on the labyrinthine story of electric power development and the contributions of those unknown “radicals” who believed in ac distribution for all loads even in cities with dense load concentration.
For Further Reading
“Alternating current supply in New York City: I,” Elect. World, vol. 63, no. 6, pp. 307–311, Feb. 1914.
“Alternating current supply in New York City: II,” Elect. World, vol. 63, no. 7, pp. 365–369, Feb. 1914.
“A compact alternating-current city substation,” Elect. World, vol. 67, no. 6, pp. 313–317, Feb. 1916.
R. W. Lobenstein and C. L. Sulzberger, “Eyewitness to dc history: The first and last days of dc service in New York City,” IEEE Power Energy Mag., vol. 6, no. 3, pp. 84–90, May–June 2008.
A. H. Kehoe, ”Underground alternating-current network distribution for central station systems,” AIEE Trans., vol. 43, pp. 844–853, discussion pp. 869–875, June 1924.
H. Richter, “Evolution of the ac network system,” Elect. J., vol. 22, pp. 320–336, July 1925.
F. W. Smith, “Development of an alternating current system,” Elect. World, vol. 80, no. 11, pp. 555–557, Sept. 1922.
“Changing from 7,800 to 13,200 volts,” Elect. World, vol. 81, no. 13, pp. 755–756, Mar. 1923.
F. W. Smith, “Alternating current in New York,” Elect. World, vol. 81, no. 21, pp. 1196–1198, May 1923.
J. J. Cunningham, “Architect of power: Thomas E. Murray and New York’s electrical system,” IEEE Power Energy Mag., vol. 10, no. 2, pp. 80–94, Mar.–Apr. 2012.