01Executive summary
Five milestones that removed the distance limit on electricity supply, decided a commercial contest fought partly by public smear, and ended in the two systems being combined rather than one winning.
The Gaulard-Gibbs transformer system was demonstrated in London in 1881 and acquired for America by Westinghouse in 1885. Stanley installed a working alternating-current distribution system at Great Barrington in 1886. Tesla was awarded the polyphase patents that gave alternating current a usable motor. Lauffen to Frankfurt in 1891 transmitted at 30,000 volts over 175 kilometres. And from 1892 the rotary converter let alternating-current transmission feed direct-current users.
02The physical argument, stated plainly
Electric power is the product of voltage and current. Losses in transmission are proportional to the square of current. So the same power can be delivered with far less loss by raising voltage and lowering current in proportion — double the voltage and losses fall to a quarter; raise it tenfold and they fall to a hundredth.
That is not a preference, it is arithmetic, and it decides everything. The question is only whether you can change voltage. Direct current cannot be transformed, so distribution voltage and utilisation voltage are the same number, and it must be low enough to be safe at a lamp. Alternating current can be transformed to any voltage a transformer is wound for. Step up at the plant, transmit, step down at the customer.
Faraday's 1831 result was that a current is induced only while a magnetic field is changing. A transformer therefore cannot work on steady direct current at all — there is nothing changing to induce anything in the secondary. Alternating current changes continuously by its nature, so it transforms freely. The entire architecture of every power system in the world follows from that one qualifying word in a laboratory result from half a century earlier, and it is the cleanest available illustration of what this series is arguing about applied science.
03The contest, and how it was fought
Alternating-current systems were being developed in Europe through the 1880s, the Gaulard-Gibbs system of Paris being among the most successful and first demonstrated in London in 1881. Westinghouse acquired the American rights in 1885 and immediately set his engineering staff to improving the generators and particularly the transformers. Stanley, appointed chief engineer that year, devised an efficient distribution system and installed it at his own expense at Great Barrington, Massachusetts, serving about 150 lamps for streets and stores. Westinghouse installed a commercial plant at Buffalo in November 1886.
Within a year the alternating-current system was a serious competitor to the established direct-current one. The response of the direct-current operators, who had held a virtual monopoly, was not primarily technical.
From the middle of 1888 the direct-current interests attempted to discredit alternating current with material purporting to show it was a horrible menace to human life, including public demonstrations of its lethality on animals. The attacks became very bitter and the two systems developed separately for a period. This is worth recording accurately in an engineering history for two reasons. First, there is a real underlying technical question about the relative hazard of the two at given voltages, and it was not the question being argued. Second, a commercially motivated safety campaign is difficult to distinguish from a legitimate one at the time, and the profession's answer to that is evidence and independent assessment rather than rhetoric. The episode is a cautionary case, not an amusing anecdote.
The separation was for a time rational rather than merely factional. Direct current generally served urban districts where demand was concentrated in a small area, which is exactly where its distance limitation does not bite. Alternating current served outlying territory where transmission lines were necessarily long. Each was being used where its physics suited.
04The missing piece: a motor
The early Westinghouse installations produced single-phase power at 133⅓ cycles per second, at transmission voltages nominally of 1,000 or 2,000 volts, and they provided electricity for illumination only. A reliable alternating-current motor did not exist until about eighteen months after the Buffalo installation.
That is a serious commercial limitation. Lighting load is concentrated in the evening; industrial motor load runs through the day. A system that can only sell light has poor plant utilisation and a much weaker economic case.
Polyphase
Generator coils wound to produce two or more separate circuits whose alternations follow in sequence as the machine rotates. The sequence produces a rotating magnetic field, which is what an induction motor needs to start and run.
Several developed it; one held the patents
A number of people in the United States and Europe developed polyphase generators and motors independently in the late 1880s. Tesla was awarded the patent rights after much litigation. Both halves of that sentence are needed.
Frequency: a standard settled by two competing requirements
The early 133⅓-cycle alternators were too high in frequency for efficient motor operation, so 25 and 60 cycles gradually became standard in the United States. The choice between them was made on opposite grounds. At 60 cycles the eye does not detect flicker in an incandescent lamp; at 25 cycles flicker is quite noticeable, particularly in small lamps. But lower frequency suits motors. So 60 cycles was used where lighting load predominated and 25 where motor power mattered more — and elsewhere 50 cycles became the lighting standard and 16⅔ was used for traction.
This is a good example of a standard emerging from a conflict rather than from an optimum. There is no frequency that is best for both lamps and motors; the settled value is a compromise whose position depends on the load mix at the time it was fixed, and which is then impossible to change because everything is built to it. Most long-lived standards have this shape.
05Neither system won
The interesting outcome is that the contest was resolved by combination rather than by victory. Before long it became evident that the advantages of each could be had together: alternating current for transmission from the power plant to outlying substations, converted there to direct current for local distribution to users who needed it.
| Device | Date | Principle |
|---|---|---|
| Synchronous rotary converter | c.1892 | A single rotating unit taking alternating current at one end and delivering direct current at the other |
| Mercury-vapour rectifier | 1902 | Hewitt's evacuated tank in which current flows from metal anodes to a mercury cathode but not the reverse — a check valve for electricity |
Direct current remained necessary for railway traction, electroplating and other industries that could use nothing else, so conversion was not a transitional expedient but a permanent part of the architecture. The modern equivalent is the same: high-voltage direct-current transmission links and inverter-connected generation sit inside an alternating-current system, and the conversion equipment is a designed component rather than a compromise.
What removing the distance limit made possible
The first American hydroelectric transmission was installed by Westinghouse in 1891, carrying power from Willamette Falls about 21 kilometres to Portland, stepped down from 3,300 volts to 1,100 for city distribution and again to 50 or 100 volts for lamps, which were not yet standardised. The Lauffen to Frankfurt line the same year carried 100 horsepower at 30,000 volts over 175 kilometres and was a convincing public demonstration.
The consequence was that water-power sites too remote or too unreliable to serve a local load became worth developing, because they could feed a system in which steam plant provided standby capacity when flow was low. Mixed hydro and steam systems appear through the 1920s, and interconnected grids follow — which is where the second series in this set takes up the story.
06Takeaways for current practice
- Find the term with the square in it. Losses proportional to current squared determine the entire architecture of power systems.
- A system that serves one load type has poor utilisation. AC could not sell motor load for eighteen months and it nearly cost the contest.
- Standards emerge from conflicts, not optima. No frequency suits both lamps and motors, and the settled value reflects the load mix of the day.
- Combination beats victory more often than the story suggests. Conversion equipment is permanent architecture, not a transitional compromise.
- Treat a commercially motivated safety argument with the same evidence standard as any other. The 1888 campaign is the cautionary case.
Modern references include AS/NZS 3000 for wiring, AS 2067 for high-voltage installations and AS 60076 for power transformers. Cited by number for orientation only — verify currency.