Readers are encouraged to share their views on issues affecting the electric power engineering profession. Send your letters to Mel Olken, editor in chief, firstname.lastname@example.org. Letters may be edited for publication.
On Matters of History
This message is for Mr. Blalock, the author of many of your interesting articles on matters of historical interest. I am a retired electrical engineer with a power system background, living in Niagara on Lake Ontario, just across the river from Fort Niagara in New York. The Niagara region of both New York and Ontario is home to numerous places of electrical historical interest. One such is the DeCew Falls hydroelectric plant that was named an IEEE Milestone in Electrical Engineering in May 2004, the Milestone nomination having been made by the IEEE Hamilton, Ontario, Section. DeCew was first studied in about 1894, with Nicola Tesla as one of the advisers. Construction started in 1896, and first power was in 1898. The objective was to send power to the steel mills in Hamilton, about 35 mi distant. The first four machines were quite small; from memory, I think they were only 100 kW each. I forget the machine voltage, but the transmission voltage was 22.5 kV, which implied a test voltage of 45 kV, which was quite a stretch in 1896. Frequency was “4,000 alternations per minute,” which we take to mean 66-2/3 Hz, although some people say this means only half that, or 33-1/3 HZ.
Tesla being what he was, the machines were two phase, as were some of the machines in the Berkshires. And this brings me to the nub of my question. It has always been my view that to produce a rotating magnetic field with only two phases, the phases had to be at 90° displacement as 180° displacement would result in a phase cancellation. But at 90°, the two phases would have been seriously out of balance. According to an article published at the time of commissioning DeCew in 1898, the four machines, which were horizontal axis type, were mounted on heavy timbers “to ensure smooth operation.” This surely would indicate that serious shaking had been experienced. Subsequent machines, built in 1906 and still in service, are all three phase. The original four machines were cut up and scrapped during World War II.
To compound the problem, I also have a photograph of a machine taken at Shawinigan Falls, Quebec, in 1901. The caption says it was 10 MW (which was really large in those days) 30-Hz, two phase. Surely that also would have shaken really badly, even if heavily constructed and firmly mounted on concrete.
If Mr. Blalock has any comments on all this, I would be most happy to hear them.
–David E. Hepburn
Mr. Hepburn has correctly identified the peculiar nature of the two-phase system in that it is, inherently, a nonsymmetrical (or “asymmetrical”) system of power distribution.
First of all, however, a clarification of “4,000 alternations per minute” is needed. This value does correspond to a frequency of 33 1/3 Hz (cycles per second) because there are two “alternations” in each cycle, one positive and one negative.
It is possible, of course, to design an alternator for any number of phases simply by providing the necessary number of individual windings at the appropriate phase displacement to each other: three windings at 120° for three phase, four windings at 90° for four phase, six windings at 60° for six-phase, and so on (I have never encountered any case where somebody operated a five-phase system, but it would be possible).
The two-phase system was, in fact, one-half of a four-phase configuration. However, the asymmetry was in regard to the nature of the two currents external to the two-phase generators or motors. Within these machines, two independent windings were evenly distributed around the periphery of the stator just as three independent windings are evenly distributed around the stator of a three-phase machine.
Therefore, there is no reason for the two-phase currents in these windings to produce any sort of significant mechanical or magnetic vibration in the machine.
The example of the early generators being mounted on heavy timbers “to ensure smooth operation” would, no doubt, have reflected concerns about the possibility of vibrations in the turbines rather than in the alternators.
Correction to “History” Column
My “History” column article, “The Schoellkopf Disaster: Aftermath in the Niagara River Gorge” published in the November/December 2012 issue of IEEE Power & Energy Magazine  contains a minor error. In Figure 4, on p. 84, the identification of the two Canadian hydroelectric powerhouses on the upper Niagara River should be reversed. The Canadian Niagara Power Co. (Rankine Station) is closest to the Horseshoe Falls and the Electrical Development Co. (Toronto Power) is further upstream. It should be noted that the figure was printed as submitted; the error is mine.
 C. A. Woodworth, “The Schoellkopf disaster,” IEEE Power Eng. Mag., vol, 10, no. 6, pp. 80–96, Nov./Dec. 2012.
–Craig A. Woodworth