|
|
|
HISTORY The Frequency Changer EraInterconnecting systems of varying cycles |
| Thomas J. Blalock | ||
|
This article describes the development of the rotating frequency changer and its application, over the past century, in a variety of utility, industrial, and transportation power systems. The Early YearsDuring the early 1890s, the Westinghouse Company attempted to introduce a dual-frequency standard for alternating current (ac) generation. The use of 60 cycles (60 Hz) was proposed for general power distribution because it was low enough to allow the use of the new Tesla induction motor but high enough to keep the required size of transformer cores reasonably small. For situations involving the use of large slow-speed machines, such as alternators driven by reciprocating steam engines or very large industrial motors, it was proposed that a frequency of 30 cycles be adopted. The possibility of interconnecting systems operating at these two frequencies may have had some influence on the selection of this exact two-to-one frequency ratio.The standard of 60 cycles, of course, has persisted until the present time. However, the success of the first ac powerhouse at Niagara Falls in 1895 established the use of 25 cycles as a low frequency because it was employed for that pioneering installation. A relatively small number of 30-cycle power systems were eventually placed into operation, primarily in the north central portion of the United States and in Canada. For example, by 1900, there was a 25,000-kW, 30-cycle hydroelectric installation at Shawinigan Falls in Quebec, Canada. This had become a significant site for the manufacture of aluminum, and the low frequency of 30 cycles was used to operate large rotary converters, which produced the tremendous amounts of direct current power required for that process.
During the late 1890s, General Electric attempted to introduce 40 cycles as a compromise to the dual frequencies of 25 cycles and 60 cycles. The latter two frequencies were too well established by that time, however, for this to succeed. Regardless, a substantial number of 40-cycle installations were made in the northeastern section of the United States. An 1898 vintage hydroelectric station on the Champlain Canal north of Albany, New York, was still using its original 40-cycle alternators through the 1980s! Also, in the late 1890s, the use of 50-cycle power was begun in connection with several independent hydroelectric installations in the mountains of southern California. Apparently, one reason for this was the fact that the General Electric Company had designs for 50-cycle alternators, at that time, which were intended primarily for use in European markets. Thus, at the start of the 20th century, the five different frequencies of 25, 30, 40, 50, and 60 cycles were in use, to a greater or lesser degree, in the United States. There was also a smattering of other oddball frequencies. The Introduction of the Rotating Frequency ChangerAs a result of the number of frequencies in use, the concept of the rotating "frequency changer" was developed early in the last century in order to be able to interconnect systems operating at two different frequencies. The first, and most straightforward, device to perform this function consisted of two synchronous machines that were coupled together. The most common arrangement was a 25-cycle machine and a 60-cycle machine, because these had become the two frequencies that were most widely used. Either machine could function as a motor driving the other machine as a generator. Thus, a reversible power interconnection could be established between two systems operating at these frequencies.As these were both synchronous machines, the frequency ratio between the two interconnected power systems had to remain absolutely constant. The unfortunate odd ratio of 25 cycles to 60 cycles caused complications when starting and synchronizing a frequency changer set between systems operating at these frequencies. A special procedure termed "pole-slipping" had to be resorted to in order to achieve synchronism to both power systems, and specially marked synchroscopes were used to aid in this task. This procedure consisted of reversing the field polarity for either or both machines. The standard design for 60-cycle to 25-cycle frequency changers used a 24-pole machine on the 60-cycle end and a 10-pole machine on the 25-cycle end. Thus, a field reversal on the 60-cycle machine caused a 75-degree phase shift on the 25-cycle end [180 x (10/24)], and a sufficient number of reversals could bring the 25-cycle machine into synchronism with the 25-cycle power system.
Also, the only way in which the division of power between two such sets operating in parallel could be controlled was to alter the phase angle through one of the sets. This was accomplished by providing a mechanism to allow the frame of one machine (of one set) to be physically rotated by a number of degrees comparable to one pole pitch of the machine. This feature could also be used during starting, in conjunction with the pole-slipping procedure. Eventually, the 1898 40-cycle hydroelectric station referred to above was equipped with two such synchronous frequency changers, which allowed the 40-cycle alternators to feed into the 60-cycle local power grid. Neither of these sets, however, was equipped with a frame-shifting device; thus, the two sets were never paralleled on the 40-cycle side. Two electrically isolated groups of generators fed these two frequency changer sets independently. In 1905, a 30-cycle to 60-cycle frequency changer of 5,750-kW capacity was installed by the Allis-Chalmers Company for the Shawinigan Water and Power Company at the Shawinigan Falls complex referred to above. Similarly, during the first half of the 20th century, several 50-cycle to 60-cycle frequency changers were installed by the Southern California Edison Company, which had inherited the 50-cycle frequency from the very early hydroelectric installations in southern California. Some of these machines were in use until the company was able to complete a conversion to 60-cycle generation late in the 1940s. In fact, in the original design for the powerhouses at Hoover Dam (also known as Boulder Dam) in the 1930s, four of the 17 generators there were to provide 50-cycle power for Southern California Edison. When the final conversion to 60 cycles was undertaken, a large expense involved the modification of customers' electric clocks; otherwise, they would have gained 12 minutes each hour at the higher frequency! As power systems grew in size and complexity, the need arose for a means to interconnect systems operating at two different frequencies in such a way as to provide a voltage tie as well as a power tie. The straight synchronous type of frequency changer provides no such voltage tie because the only actual connection between the two systems is mechanical via the coupled shafts. The Induction-Synchronous Frequency ChangerThus, during the 1920s, the induction-synchronous type of frequency changer was developed. This consisted of a synchronous machine coupled to a slip-ring (wound rotor) type of induction motor. The synchronous machine operated from one frequency (usually the lower of the two), and this same frequency was also fed to the rotor of the induction machine. The stator of the induction machine operated from the other (higher) frequency system. This induction machine acted as a "rotating transformer," providing a voltage tie between the two systems. The power flow (in either direction) from one system to the other took place partially through the connecting shaft and partially through the air gap of the induction machine. Because of its ability to transform frequency as well as voltage (and the number of phases, if desired), this type of machine was referred to as being a "generalized alternating current transformer" by Dr. Charles P. Steinmetz, the renowned General Electric scientist of that period.In New York City, such induction-synchronous frequency changers were installed early in the 20th century to couple the 25-cycle system of the New York Edison Company with the growing 60-cycle system of the United Electric Light and Power Company. The former system had developed primarily as a result of the use of rotary converters to feed the entrenched Edison direct current (dc) distribution system in Manhattan. The early rotary converters were prone to flashovers if it was attempted to operate them on frequencies as high as 60 cycles. These frequency changers were in use until the late 1970s, when the rotary converters were finally retired and the remaining dc load was fed by means of solid-state rectifiers. The first of these frequency changers was installed at the huge Hell Gate Generating Station, in the Bronx, New York City, in 1923. This was a 35,000-kW machine that was not equipped with a frame-shifting device to control power flow because it was the only frequency changer coupling the two systems. Later, however, 40,000-kW induction-synchronous sets were installed at both the Waterside and East River Generating Stations in Manhattan, New York City. As these machines operated in parallel with the Hell Gate frequency changer, they were all equipped with frame-shifters in order to control the division of power flow through the various machines. By the 1930s, the induction-synchronous type of frequency changer had been developed to include a variation known as a "Scherbius" frequency changer. This machine utilized a complex speed control system, which had been developed originally for use with large slow-speed induction motors. This speed control feature allowed for flexibility in the frequency ratio between the two power systems, which were interconnected, in order to be able to accommodate slight variations in frequency on either system. Thus, the term "variable-ratio" was applied to frequency changers that used the Scherbius control system. Frequency Changer Installations: the End of an EraDuring the 1930s, a total of three 20,000-kW Scherbius frequency changers were installed at St. Louis, Missouri, to convert 25-cycle power from a huge 1913 vintage hydroelectric station located on the Mississippi River at Keokuk, Iowa, into 60-cycle power. These frequency changers were still in use in the mid-1980s.
The steel industry also embraced 25-cycle power during the early decades of the 20th century. This had to do with the use of huge slow-speed gas engines (operating on blast furnace waste gas) to generate power, as well as the need for large slow-speed induction motors as rolling mill drives. Slow-speed 60-cycle motors having a large number of poles operated at rather low-power factors. Motors operating at 25-cycles were preferable for these types of applications, and the development of suitable heavy-duty reduction gears for use in rolling mill drives did not occur until later in the 20th century. In 1935, a 20,000-kW outdoor, hydrogen-cooled frequency changer of the straight synchronous variety was installed to interconnect the 25-cycle power system of the former Bethlehem (Pennsylvania) Works of the Bethlehem Steel Company with the 60-cycle power system of the Pennsylvania Power and Light Company. That frequency changer was in use through the mid-1990s and has only very recently been scrapped. Meanwhile, a second 20,000-kW frequency changer was installed during the 1940s to operate in parallel with the existing frequency changer. This latter unit was a more conventional indoor set and was equipped with a frame-shifting mechanism on its 60-cycle machine in order to control the division of power flow through the two sets. This indoor frequency changer was in use until just a month or so ago in order to supply 25-cycle power to some old motors in buildings that are now being used for other purposes. In the Niagara Falls region, a total of five frequency changers were in use by the Niagara Mohawk Power Company during the 20th century. There was one Scherbius unit at the Huntley Generating Station in Buffalo, New York, and two Scherbius units at Lockport, New York. In addition, there were two units located in Gardenville, New York, a suburb of Buffalo. These latter units were not Scherbius and, in fact, had been purchased from the Consolidated Edison Company of New York City in 1957. They were still in operation in 1999. All of these machines are now out of service, however, and the relatively few remaining 25-cycle customers in western New York are now being supplied with power imported from Canada. Present plans are for that service to be terminated by the year 2007, according to Craig Woodworth, a retired Niagara Mohawk engineer from Tonawanda, New York, who is very well informed on the subject of 25-cycle power in western New York. Single-phase power at 25 cycles is still being used to operate electrified trains on the Amtrak main line from New York City to Washington, DC. This use of 25-cycle power originated with the electrification of this line in the 1930s by the former Pennsylvania Railroad Company. The series-type ac traction motors could not be operated on frequencies as high as 60 cycles. Some of this 25-cycle power is still being supplied from 25-cycle generators and a frequency changer, located at the Safe Harbor Dam on the Susquehanna River in Pennsylvania. Also, at Metuchen, New Jersey, a 1930s vintage frequency changer is being maintained (although parts are becoming hard to find) by the Public Service Electric and Gas Company. This is a 25,000-kW synchronous-synchronous machine that was built by the General Electric Company. It is equipped with a frame-shifting device to allow it to operate in parallel with 25-cycle sources at Safe Harbor as well as in New York City. In New York City, however, previous rotating sources of 25-cycle power have been replaced by solid-state "cyclo-converter" units. This trend is an indication that the 100-year era of the rather fascinating rotating frequency changers will soon be over. For Further Reading"The MechanicvilleSchenectady Power Transmission Plant," Electrical World, pp. 233-236, Sept. 1898.Shawinigan, "The Electrical Age," Electric J., pp. 357-360, 1907. B.G. Lamme, "The technical story of the frequencies," AIEE Trans., pp. 65-89, Jan. 1918. W.J. Foster, "Early days in alternator design," General Electric Rev., pp. 80-90, Feb. 1920. "A 35,000-kilowatt induction frequency converter," AIEE Trans., pp. 1011-1024, June 1924. "Automatic control for variable-ratio frequency converters," AIEE Trans., pp. 121-127, Mar. 1932. "Developments in the electrical industry during 1935," General Electric Rev., pp. 7-8, Jan. 1936. "Switchboards for Boulder Power Plant," AIEE Trans., pp. 224-236, Feb. 1937. P. Mixon, "Technical origins of 60-Hz as the standard AC frequency in North America," IEEE Power Eng. Rev., pp. 35-37, Mar. 1999. "The Y1936 problem," Invention Technol., p. 64, Winter 2000. |