Page images
PDF
EPUB

The most serious danger, in the growth of electric systems, was, however, the possibility of self-destruction by the power let loose under short circuit, and there were anxious years for the operators and managers of these large electric systems before the industry devised the means of safely controlling unlimited power. More than once, when a serious short circuit occurred and a disaster was averted only by luck, the system was cut into two or three sections and these operated independently, to limit the power. But when months passed without further accidents, invariably, due to the requirements of economy and reliability of operation, the sections came together again and parallel operation of the entire system was restored.

This, the most serious problem of the high-power electric system, was solved by the development of the power limiting

reactances.

In the generator leads, between generators and busbars, are inserted reactances, capable of standing enormous overloads, of a size sufficiently small not to interfere with the normal flow of power, at full load or any overload which the generator may be called upon, but large enough to materially limit the generator current and power at short circuit. Usually the generator reactances limit the momentary short-circuit current to about ten to twelve times full-load current; that is, the momentary shortcircuit power to about two and one-half times full-load power. This solved the problem for medium-sized stations. Thus in a 60,000-horse-power station, instead of a possible short-circuit power of over half a million horse-power, the power is limited to 150,000 horse-power.

However, even with generator reactances, with increasing size of station, the power which may be let loose under short circuit becomes large beyond control; with a 400,000-horse-power station, with generator power limiting reactances, a million horsepower may still be concentrated at a short circuit.

Busbar reactances then were introduced; that is, the busbars divided into sections by reactances sufficiently small not to interfere with the interchange of power along the busbars, and thereby retaining the advantage of parallel operation, but large enough to limit the flow of power, which, in case of a short circuit on one busbar section, can flow into it from the adjoining sections.

By such reactances in the busbars and in the tie feeders be

tween the stations, the system is divided into sections of about 60,000 horse-power each. A short circuit then can seriously involve one busbar section only, and the destructive power of a short circuit is limited to that of one section, plus the limited power which can flow from the two adjoining sections, a total of 150,000 to 200,000 horse-power, and this is within the emergency limit of the modern oil circuit breaker of moderate size. It still represents a terrific energy, nearly 10,000,000 foot-pounds, and is a severe strain on the circuit breaker.

These busbar reactances permit an unlimited extension of the system, and the short circuit on a section of a half-million-horsepower system is no more severe than a short circuit on a 100,000horse-power system, and there is now no limitation to the future. increase to electric systems of many millions of horse-power capacity, operating in parallel on one set of busbars.

With hundreds of feeders radiating from the busbars, the probability of a short circuit in feeders is far greater than in the busbars, and a material advantage, therefore, is given by feeder reactances; that is, reactance interposed between the feeder or a group of feeders and the busbars, so that a short circuit in the feeder is still more limited than a short circuit in the busbars.

By the development of generator reactances, busbar reactances, and feeder reactance, the problem of the power control of large systems for protection against self-destruction by short circuit has been solved and unlimited extension of systems without any increase of danger has been made possible, and experience has shown that after the introduction of such power-limiting reactances dead short circuits have occurred at the busbars of very large systems without even interfering with the operation. of most of the synchronous apparatus on the system.

Not all three classes of reactances are always necessary in systems of moderate size busbar reactances may not yet be needed. In low-head water-power plants, with slow-speed multipolar alternators of inherently limited short-circuit current, generator reactances may be unnecessary and only busbar reactances required. Such, for instance, is the case at the Keokuk plant on the Mississippi River. Again, with a perfect system of generator and busbar reactances, feeder reactances may be dispensed with— though they are even then an advantage.

In high-voltage transmission networks, even of very high

power, power-limiting reactances sometimes may not be required or less essential. With a considerable number of medium-sized water-power plants feeding into a transmission system, the power of each individual generating station may not be sufficient to give destructive values under short circuit, and the impedance of the lines between the generating stations may be sufficient to limit the power which can feed into the short circuit at one station. In transmission networks, therefore, power-limiting reactances are necessary only in very large generating stations, as the Keokuk station, or where several fairly large stations are close together, and also, as generator reactances, in turbo-alternators connected into the system as steam reserve.

To cut off a disabled line or feeder with the voltages and powers of our modern systems is beyond the capacity of the fuse or simple blade switch, and aut matic oil circuit breakers are generally used. However, the problem has become more difficult by the increasing demand for reliability and continuity of service.

The two main sources of troubles in lines and cables are grounds and short circuits between phases. In transmission lines a ground on one phase is the most frequent trouble, and short circuits are rare except in lines in which the design was faulty, or reliability had been sacrificed to cheapness, and the spacing between conductors chosen too small, so that they swing together during wind storms, etc. A short circuit is far more serious than a ground, as in the former the current is limited only by the generator capacity, while with a ground the current has no return— except if the neutral is grounded, and then over the resistance of the neutral-and the current, and with it the shock on the system, therefore is very much less, especially if safeguards against the occurrence of high frequency by arcing grounds are installed. In a well-designed transmission line a short circuit usually occurs only as the result of two simultaneous grounds. A ground on one conductor, however, raises the voltage against ground, of the other two phases, from the Y voltage to the delta voltage of the system, and thereby increases the strain on the insulation of the other two phases. It thus either introduces the danger of a second ground, causing a short circuit, or requires a higher grade of insulation.

This has led to two methods of operation of transmission. systems. Either the neutral of the transformers is grounded, VOL. CLXXX, No. 1075-2

frequently through a resistance, where the resistance of the ground is not high enough to limit the current. Then a ground on one phase is a partial short circuit to the neutral, causes a large current to flow, and thereby opens the automatic circuit breakers and cuts off the circuit before the ground has developed to a short circuit. However, this method, the " grounded Y system," means a shutdown at every ground, every flashover of an insulator by lightning, etc. Or the neutral of the system is not grounded, the insulation of the circuit made good enough to safely stand the increased strain put on it by a ground on one phase, and by an arcing ground suppressor, etc., care taken not to continue an arcing ground-leading to high-frequency disturbances—but convert it into a metallic ground. In this case, the "isolated delta" system, service can be maintained on the circuit, even if one phase grounds, until arrangements are made to take care of the load, or the fault found and remedied, and the continuity of service thus is not interfered with. However, the cost of line construction is higher, due to the better insulation required. The relation between grounded Y and isolated delta thus is that of cheapness versus reliability and continuity of operation, and, as a rule, we find grounded Y systems where lowest cost of development is considered essential and occasional interruption of service not considered objectionable, while the isolated delta is generally preferred in systems in which reliability and continuity of service are considered as of first importance, such as in the extension across the country of the great Metropolitan Edison Systems— systems which are proud of their record that the voltage has not been off their busbars for ten years or more.

Different are the conditions in underground cable systems. In a cable the three conductors are so close together that a ground on one conductor quickly reaches the other conductors and becomes a short circuit. A grounded cable, therefore, cannot be kept in service, but has to be cut out as promptly as possible. In these systems it therefore is customary to ground the neutral through a resistance sufficiently low, in case of a ground on one conductor of a cable, to allow sufficient current to flow to open the circuit breaker and cut off the cable, but sufficiently high not to give a severe shock on the system. Or, where grounding of the neutral is considered undesirable, an arrangement of relays is made to give the same effect. With underground cables such

cutting off of a disabled feeder does not interfere with the continuity of service, as a number of feeder cables are always used in multiple for every important substation.

However, the problem of cutting off a disabled feeder by the operation of the circuit breaker, due to the large current taken by the grounded feeder, is not so simple. Assuming that three cables feed in multiple into a substation, and one of these feeders grounds: a large current then flows from the generating station into this cable to ground, and the circuit breaker at the generator end of this feeder opens. This, however, does not stop the current rush, but a large current still flows through the damaged cables into the ground, coming back from the substation, and flows to the substation from the generating station through the two parallel feeders, which are undamaged; that is, short-circuit current feeds back through these two cables over the substation, and these two cables also open their overload circuit breakers, cutting off and thereby shutting down the substation. If the substation is connected by tie feeders to adjoining substations, current feeds back into the faulty cable over these tie feeders from adjoining substations, and these tie feeders, and the cables feeding the adjoining substations from the generating station, open their circuit breakers by overload, and in this manner a ground in one cable may shut down a number of substations, possibly the entire system. Time-limit devices in the circuit breakers are insufficient to protect against such extended shutdowns resulting from a single fault in a cable. A permanent time limit is not permissible in large systems, as with a dead short circuit the circuit breakers must open instantly before extensive damage is done by the large power of the short circuit. Therefore so-called “inverse time-limit" circuit breakers are generally used; that is, circuit breakers in which the time limit of their operation decreases with increasing overload. Such circuit breakers would first cut off the cable carrying the greatest excess current, that is, the faulty cable,—and then those of less excess current; but, as with the cutting off of the faulty cable-at both ends-the excess current stops, other cables should not be interfered with. However, the inverse time-limit circuit breaker necessarily must be practically instantaneous under short circuit, and therefore, while the time limit discriminates between 100 per cent. or 200 per cent. or 300 per cent. overload, it cannot discriminate between

« ՆախորդըՇարունակել »