history

Thomas J. Blalock

the first polyphase system

a look back at two-phase power for ac distribution

THIS ARTICLE DESCRIBES THE

first polyphase (more than one
phase) system developed for the distribution of alternating current (ac)
power. This two-phase system was
subsequently rendered obsolete, however, by the superior three-phase system that is now universally used
throughout the world.

The Origins of
Two-Phase Power
Today, the large-scale generation, transmission, and distribution of electric
power is by means of the three-phase
ac system; that is, three individual single-phase voltages and currents having
a 120 phase relationship to each other
and intermingled on three wires
(excluding a neutral). The three-phase
system has been adopted because it
provides for a constant rather than pulsating power flow to motors, and
because it is an efficient system as far
as the amount of copper required per
kilowatt transmitted. The theoretical
complexity of the three-phase system,
however, delayed its complete acceptance in the early days of electric power
system development.
During the early 1890s, understanding the behavior of simple single-phase
ac was enough of a challenge. It was
not until Charles P. Steinmetz, the legendary General Electric scientist,
developed the concept of the use of the
j operator (unity magnitude at a 90
phase angle) and complex numbers for
ac circuit calculations that the behavior
of voltages and currents in ac circuits
march/april 2004

and machines was truly understandable. Likewise, it was not until the
introduction of what eventually came
to be known as symmetrical components, during the early 20th century,
that the calculation of three-phase voltages and currents became relatively
straightforward. This technique utilized
an a operator that was of unity magnitude at a 120 phase angle (0.5 +
j0.866). This operator was of significant value since, in a balanced threephase system, the voltages and currents
are at 120 phase relationships to each
other.
Symmetrical components actually
facilitated calculations in unbalanced
three-phase circuits. They were originally known as Fortescue components since the method was introduced
in 1918 by Charles L. Fortescue of the
Westinghouse Electric Corporation.
Significant additional work in this area
was later contributed by Edith L.
Clarke of the General Electric Company. During the late 19th century, however, this calculation tool did not exist,
and the fact that changes in voltage or
current magnitudes in one phase of a
three-phase system affected the voltages and currents in the other two phases contributed to the difficulty in
understanding three-phase circuits.
Thus, the first ventures into the
realm of polyphase electric power used
only two alternating current phases
rather than three. The two phases were
generated with a 90 phase difference
between them, and the system that
resulted was called two-phase power.
ISSN 1540-7977/04/$17.002004 IEEE

In fact, the first two-phase generators

employed during the early 1890s were
merely two single-phase machines coupled together with their rotors carefully
set relative to each other so as to
achieve the required quadrature phase
relationship. Each generator, then, really fed a separate two-wire, single-phase
circuit. Since the two phases were completely electrically isolated from each
other, there were no interactions
between voltage and current magnitudes in one phase with those quantities
in the other phase. Therefore, from a
theoretical standpoint, the two-phase
system was more easily understood
than was the three-phase system.
The two phases were used together
in a four-wire system to enable the
operation of the new Tesla (or induction) motor that had been developed
by Nikola Tesla. In order to be selfstarting, the Tesla motor required
some form of rotating magnetic field
that had to be produced by a
polyphase type of supply. The twophase system was adequate for this
purpose. The Westinghouse Electric
Corporation supplied the power plant
and lighting for the Colombian Exposition in Chicago in 1893. Two-phase
power, produced by pairs of coupled
single-phase generators, was used
throughout this installation.

Two-Phase Power
at Niagara Falls
The experience gained with the use
of two-phase power at the Colombian
Exposition may have had some
IEEE power & energy magazine

63

Original two-phase to three-phase transformers installed at Niagara Falls in

1895 (photo courtesy Hall of Electrical History at the Schenectady Museum,
Schenectady, New York).

influence on the decision by Westinghouse to employ a two-phase generator

design for the first ac powerhouse at
Niagara Falls, which went into operation in 1895. The generators used at
Niagara Falls were of a more conventional design, being single machines
having two interleaved windings rather
than two distinct machines coupled
together. These generators operated at
a frequency of 25 cycles (25 Hz) since
it was expected that a significant portion of the power produced would be
used to operate rotary converters so as
to obtain direct current (dc) for industrial uses such as aluminum production. These early rotary converters
required a low frequency for satisfactory operation.
There was obviously still a mistrust
of the practicality of three-phase
power throughout the electric power
industry at that time. For example,
according to an 1896 article titled
Present Status of the Transmission
and Distribution of Electrical Energy
in the AIEE Transactions:
Where a two-phase transmission with separate circuits is
used, then if the separate circuits
64

IEEE power & energy magazine

are wound on different armatures, each can be regulated to

give a constant voltage at the
receiving end. This is the case,
for instance, in the large
dynamos built by the Westinghouse Company for use at the
Worlds Fair in Chicago. The difficulty due to the uneven loading
of the circuits is specially
marked in the case of the threephase system, and it is one of the
principal objections that have
been urged against the employment of this system for purposes
of distribution.
It had already been realized, however, that the three-phase configuration
was superior for transmission from the
point of view of efficiency. Thus, special phase-changing transformers were
designed by Charles F. Scott of Westinghouse in order to step up the twophase generated voltage at Niagara
Falls to 11,000-V, three-phase for
transmission to Buffalo, New York.
The General Electric Company was
awarded the contract to build the
phase-changing transformers and so
was licensed by Westinghouse to uti-

lize the connection developed by Scott

for this purpose.
At Buffalo, some of this three-phase
power was used for rotary converters
that supplied 110/220-V dc power for
the Edison distribution system downtown. However, some of the received
power was converted back into twophase power for general lighting purposes in outlying areas. Motorgenerator sets were used for this latter
conversion because the frequency of
the ac power was increased as well in
order to avoid undesirable flickering of
incandescent lamps. The frequency
used was actually 62.5 cycles, rather
than 60 cycles, so as to simplify the
design of these frequency changers.
The conversion back to two-phase
power was motivated by the conviction,
at that time, that satisfactory voltage
regulation was more easily achieved in
the two separate phases of a two-phase
system than in a three-phase system.
This belief in the superiority of twophase systems with respect to voltage
regulation led to the extended use of
two-phase distribution in many locales.
For example, in Cohoes, New York,
(north of Albany) a 1915 hydroelectric
station was designed to generate threephase power. However, some of that
power was converted to two-phase
using Scott type transformers in
order to supply an extensive network of
existing two-phase feeders for lighting,
rather than change those feeders to
three-phase operation.

William Stanley
Adopts Two-Phase
William Stanley, the man credited
with the first practical application of
the ac system using transformers (in
Great Barrington, Massachusetts, in
1886), subsequently formed the Stanley Electric Manufacturing Company
in Pittsfield, Massachusetts, in 1891.
Stanley adhered to the design and construction of two-phase generators and
motors throughout the 1890s. This
was only partly a result of his belief in
the superiority of the two-phase system for voltage regulation purposes.
Another factor had to do with the
march/april 2004

increasing development of three-phase

equipment by his major competitors,
General Electric and Westinghouse,
during the 1890s. Stanleys decision to
manufacture two-phase equipment
allowed him to avoid excessive patent
infringement problems with his competitors. Regardless of the reasons,
however, Stanley contributed to the
perpetuation of the use of two-phase
power in many locations.
The Stanley Works itself generated
and utilized two-phase power. In 1907,
this plant became the Pittsfield Works
of the General Electric Company, and
the two-phase power system that it had
inherited from Stanley remained in use
until the closing of the facility in 1987.
In fact, to this day, there is still one elevator in an old office building there
operating with a two-phase motor.
The two-phase system in this plant
was somewhat unusual in that it was a
three-wire system. One wire from each
phase was combined into what was
called a common wire (not a neutral). The advantage in this was the
ability to use more commonly available
three-pole circuit breakers and switches.
A disadvantage, however, was that even
with the two phases balanced, the common wire carried 1.414 times the current
in the other two phase wires. Thus,
economy in pulling circuits through
conduits required the use of two different sized cables. Eventually, the plant
had two power distribution systems, the
original two-phase system and a newer
three-phase system. The two systems
were interconnected by means of phasechanging transformers. These were of a
design by Louis F. Blume of the General
Electric Company and utilized a winding configuration differing from the
Scott connection, presumably to avoid
patent conflicts with the Westinghouse
Electric Corporation.
Since Stanley supplied equipment
to the local municipal power company, the Pittsfield Electric Company,
downtown Pittsfield was also served
by a two-phase system. This, however, was the more conventional fourwire type of two-phase distribution
requiring four-pole service switches.
march/april 2004

This two-phase distribution system

remained in use until the middle of
the last century, and vestiges of it in
the form of four-pole switches could
still be found on the service switchboard of at least one old building in
Pittsfield in the early 1980s. Also,
two-phase motors were still being
used to drive the elevator motor-generator sets in Pittsfields only department store when it closed in 1988.

Other Two-Phase
Installations
In the village of Middle Falls, New
York, (northeast of Albany) the Niagara
Mohawk Power Corporation operated a
1900 vintage, 350-kW Stanley twophase generator in a small hydroelectric power station there until 1987.
Another identical unit had been retired
in 1976. The output of the station was
coupled to Niagara Mohawks threephase grid by means of phase-changing
transformers.
The generation of two-phase power
was not exclusively an East Coast phenomenon, however. In 1898, the Pacific
Light and Power Company installed
four 300-kW Westinghouse two-phase
generators in a hydroelectric station
located in San Gabriel Canyon, near
Los Angeles, California. This station
served the nearby town of Azusa.
As the use of ac motors expanded
during the early 20th century, the problem of providing both l15 V for lighting and 230 V for motor use from
two-phase distribution systems became
significant. One solution was the adoption of a two-phase, five-wire system in
which center taps on both phases were
connected together to create a neutral.
This, then, resulted in a star configuration (analogous to the three-phase
wye connection) and, technically,
was a four-phase system. As such, 115
V (single-phase) for lighting was available from any of the four phase wires
to the neutral, while 230 V (two-phase)
was available for motors from the four
phase wires themselves.
In New York City, the Bronx District of the New York Edison Company adopted this form of secondary

distribution around 1925. At that time,

the Company was interested in
upgrading its existing 2,400-V, twophase primary distribution system to
13,200 V, three-phase. The connected
two-phase motor load, however, was
too great to consider changing the
secondary distribution system from
two-phase to three-phase as well, so
T-connected (Scott) phase-changing transformer banks were installed
to supply a two-phase, five-wire secondary distribution system.
During this era, the use of the threephase, four-wire wye-connected distribution system was often considered to
be unacceptable because of the nonstandard voltage (199 V) between phases with 115 V available from phase to
neutral. Early induction motors,
designed for operation at 230 V, were
less satisfactory when operated on
lower voltages than are induction
motors of today. The ability of the twophase, five-wire distribution system to
supply the standard voltages of
115/230 V was a main feature in a
lengthy article published in the AIEE
Transactions in 1925 by an engineer
associated with the Philadelphia Electric Company in Pennsylvania. This
article justified the continued use of
that system.

William Stanleys company specialized in two-phase equipment.

IEEE power & energy magazine

65

The Demise of
Two-Phase Systems
Eventually, a hybrid type of three-phase
distribution system, which was known
as a three-phase, four-wire, delta system, came into use in certain regions of
the United States. This system included
a center tap on one phase of a bank of
delta-connected transformers supplying
230 V. The center tap formed a neutral
and, in conjunction with the two phase
wires of that particular phase, was used
to supply 115/230 V services on a single-phase, three-wire basis. Motors
operating at 230 V were supplied from
the three phase wires of this type of
service connection.
Buildings requiring both motor and
lighting service were sometimes provided with two separate services, a single-phase, three-wire service for
lighting and a three-phase, three-wire
service for motors. Otherwise, a single
four-wire service was brought into a
building, but care had to be exercised
by electricians so as not to use the odd
phase wire along with the neutral to
supply lighting loads. This odd phase
was referred to as the high phase or
wild phase because considerably

A two-phase, four-pole service switch

in a building in Pittsfield, Massachusetts (Tom Blalock photo).

66

IEEE power & energy magazine

more than 115 V existed between it

and the neutral. This complication
associated with the four-wire delta
type of service led to its gradual abandonment during the latter 20th century
because fewer and fewer practicing
electricians were able to truly understand it. Also, by that time, induction
motors had been developed that operated satisfactorily on voltages lower
than 230 V. As a result, the threephase, wye-connected service, giving
208 V between phases and 120 V from
phase to neutral, has become the standard commercial type of service. Also,
over the years, old two-phase primary
distribution systems were gradually
replaced with three-phase systems. A
common practice became the conversion of a 2,300-V, two-phase, fourwire distribution system into a
4,000/2,300-V three-phase, four-wire
system (with neutral).
Several clever and complex plans
were devised for the temporary supply
of remaining two-phase loads from a
new three-phase system, without the
expense of purchasing special phasechanging transformers. One such technique took advantage of the fact that
there is a 90 phase relationship
between one phase-to-phase voltage
and the voltage from the third phase to
neutral in a three-phase, four-wire system. Customers were encouraged to
purchase three-phase motors, rather
than add to their existing inventory of
two-phase motors. Many of the old
motors, however, lasted for quite some
time. Occasionally, a customer actually
had to be supplied with two services,
one two-phase and one three-phase.
With rare exception today, the twophase distribution system has become a
thing of the past. Its extensive use
throughout the 20th century, however,
created interesting situations for electrical engineers accustomed to threephase systems. Occasional oversights,
resulting from the unrecognized need
for four-pole motor control contactors
due to the existence of an old two-phase
system, have been known to cause

havoc for electrical equipment designers and suppliers.

For Further Reading

J.O. Kraehenbuehl and M.A. Faucett,
Circuits and Machines in Electrical
Engineering. New York: Wiley, p. 268,
1939.
Electrical Transmission and Distribution Reference Book (4th ed.). East
Pittsburgh, PA: Westinghouse Electric
Corporation, p. 12, 1950.
E.L. Clarke, Determination of voltages and currents during unbalanced
faults, General Electric Rev., pp.
511513, Nov. 1937.
C. Passer, The Electrical Manufacturers (18751900). Cambridge, MA:
Harvard, 1953.
L.B. Stillwell, The electric transmission of power from Niagara Falls,
AIEE Trans., pp. 444-486, 23 Aug.
1901.
Present status of the transmission
and distribution of electrical energy,
AIEE Trans., vol. XIII, Sept. 1896.
H.G. Stott, The distribution and
conversion of received currents, AIEE
Trans., pp. 125-163, 22 Mar. 1901.
B.R. Connell, The Hydro-Electric
Development of the Cohoes Company
at Cohoes, N.Y., General Electric Rev.,
pp. 340352, May 1915.
L.F. Blume, Transformer connections for three-phase to two-phase
transformation, General Electric Rev.
pp. 552559, Sept. 1912.
W.A. Myers, Iron Men and Copper
Wires: A Centennial History of the
Southern California Edison Company.
Glendale, CA: Trans-Anglo Books,
1983.
Distribution for congested areas,
Electr. World, pp. 1031-1032, 16 May
1925.
P.H. Chase, Two-phase, five-wire
distribution, AIEE Trans., pp.
737749, June 1925.
Changing from two-phase fourwire to three-phase four-wire distribution, Electric J., pp. 214216,
p&e
June 1923.

march/april 2004