a steady-state limit of perhaps 5.0 kv 2 for terminal reactances somewhat smaller than those used in Figure 6 of reference 3.

Pilot-Wire Circuits for Protective Relaying— Experience and Practice

Discussion and author's closure of paper 4 3 - 2 7 b y the A I E E re lay subcommi t tee of the c o m ­mi t tee on protect ive dev ices , presented at the A I E E nat ional technica l m e e t i n g , N e w Y o r k , Ε . Y . , January 2 5 - 2 9 , 1 9 4 3 , a n d pub l ished in A I E E T R A N S A C T I O N S , 1 9 4 3 , M a y sec­t i o n , pages 2 1 0 - 1 4 .

W. R. Brownlee (The Commonwealth and Southern Corporation, Jackson, Mich.) : The collection of information on experience with pilot-wire circuits for protective relay-ing should lead to a more widespread em-ployment of this form of protection in making possible marked savings in major equipment. In addition to the critical examination of the per cent availability of pilot-wire circuits, other factors affecting the over-all performance of the protective sys-tem may well be considered. Experience has shown that excellent reliability and freedom from false operations may be secured by using either privately owned or leased circuits, provided the following major requirements are met:

1. T h e relay equipment should be immune to false tripping (in the presence of load current alone) in case the pilot-wire circuit is momentarily short-circuited, open-circuited, or grounded, or in case moderate extraneous voltages are momentarily introduced. Directional-comparison schemes of pilot-wire relaying are inherently immune to such disturbances, and in some instances directional fault detectors using current and voltage have been used to interlock pilot-wire relays of the two-wire dif-ferential type .

2. Adequate supervision of the condition of the pilot-wire circuit should be provided to detect de-fects and remedy them before the relay equipment is called upon to operate. Continuous supervision has been found particularly valuable in this connec-tion, and supervision against high-resistance grounds on the circuit has resulted in detecting incipient pilot-wire short circuits or open circuits some hours (or even days) in advance of the pilot-wire circuit failure (since a single ground does not interfere with operation).

3 . Thorough and proved measures should be em-ployed to prevent momentary disturbances to pilot-wire circuits caused by rise of ground potential, induction, or other causes which by their nature occur at the same time as a power-system fault for which the relays must discriminate.

Mention is made of the ability of the "newer schemes" to provide both phase and ground protection and also to broaden appli-cation to cover greater distances. It is believed that the subcommittee intends this discrimination between early three-wire a-c pilot-wire systems and more recent develop-ments, rather than between d-c-operated directional comparison schemes which have been in general use since about 1930 and the two-wire low-energy a-c differential pilot-wire equipment appearing some seven years later. Even the most recent types of a-c two-wire differential schemes have been found to be limited to appreciably shorter lengths of pilot-wire circuits than can be used with d-c directional-comparison

schemes. In a recent application it was therefore necessary to choose d-c pilot--wire relays on account of the length and characteristics of the available pilot-wire circuits.

The results of the subcommittee investi-gation indicate some confusion in the matter of guarding against rise of ground potential and this is a very important point in the pre-vention of pilot-wire disturbances concur-rent with power-system faults. The follow-ing suggestions are offered:

1. T h e use of three-winding neutralizing trans-formers on the prlot-wire circuits is strongly recom-mended for the usual application.

2 . Extreme caution is required in segregating all portions of the circuit on the station side of the neutralizing transformer from any equipment which might be at remote ground potential and vice versa.

3 . T h e use of staged fault tests to ground is highly recommended for checking on the over-all effective-ness of voltage neutralization and discovering diffi-culties which might otherwise be obscured.

Some common causes of trouble in con-nection with rise of ground potential are :

1. Plain and simple confusion based on the mis-conception that " a ground is a g r o u n d . "

2 . Communication circuits in which a momentary interruption during a power-system fault m a y be of little consequence are sometimes not provided with neutralizing transformers or other potential gradi-ent equipment. Operation of carbon protectors on these circuits during power-system faults m a y in effect by-pass the neutralizing transformer protec-tion installed on pilot-wire circuits at the same sta-tion.

3. T h e overenthusiastic use of carbon protectors may nullify an excellent protective system.

4. Lack of thorough discussion m a y result in fail-ure to co-ordinate the protective equipment of the power company and the communication company from which circuits are leased.

5 . C o m m o n neutral distribution circuits, trans-mission-line ground wires, and so forth, m a y extend the field of station ground influence to an unsus-pected extent and thus destroy the effectiveness of remote ground installations.

H. P. Sleeper (Public Service Electric and Gas Company, Newark, N. J.) : This paper on pilot-wire relaying seems definitely in order for three reasons:

I. It has been several years since the subject has been discussed before this group.

2. This branch of the protective-relaying art is definitely rising in popularity on the list of relay-parade schedules.

3 . T h e art of pilot-wire relaying has made decided advances in the last few years, its scope of applica-tion has been extended, and its economies of appli-cation have been enhanced.

Without attempting to review or analyze the history or development of this art, it seems to this commentator that three major factors have tended to popularize the use of this type of protection. The first is the development of schemes which enable standard relays to be used with pilot-wire interconnections to obtain high-speed dif-ferential protection of transmission circuits. Examples of this are the use of directional relays to compare fault-current flows at the ends of such a circuit. Another is the use of the pilot-wire channel to transmit blocking signals between sets of standard carrier-cur-rent relays when the latter encounter tech-nical or economic difficulties in a particular application.

The second factor has been the develop-ment of a simple network type of pilot-wire relay which is small in size, economical to

apply, simple to test in service, independent of potential connections, yet supplying phase and ground protection in a single unit, and which has proved very successful in opera-tion.

A third factor is the developments which permit standard-gauge telephone circuits to be used as the pilot conductors. It is not many years ago that all relay schemes of this type required low-resistance pilot con-ductors, and such sizes as number 6 (B. & S. gauge) were commonly employed. Ob-viously, the economics were against the ex-tension of such schemes, and long-line appli-cations were not considered, the usual instal-lation being for distances of fractions of a mile. Today circuit sizes as small as number 24 (B. &S. ) are in use, and a circuit of 27 miles in length is reported in the subject paper. Much longer pilot circuits can be used, even with the small gauge telephone wires with certain relay schemes.

It is this observer's opinion that pilot-wire relay protection has a definite place in the art of protective relaying today. There arise from time to time certain applications where this general scheme is definitely supe-rior economically to any other. This is usually where short sections of a system are to be protected. The degree of protection afforded is equal to that of any of the stand-ard modern relays on such applications. On longer lines the chief competitors are dis-tance relays and carrier relays. On lines up to 25 miles, the annual costs of pilot-wire relays and carrier pilot relays are not very different.

The writer is associated with an electric utility company which has in operation pilot-wire relay protection on approximately 70 miles of three-phase transmission circuits. Two circuits totaling 10 miles operate at 26 kv, and eight circuits totaling 60 miles operate at 132 kv. Leased pilot circuits are employed throughout. On most of the longer circuits it has been the practice to install neutralizing transformers in the pilot wires, for longitudinal voltage reduc-tion. The leased circuits have had all branch tap circuits removed so as to mini-mize pilot circuit troubles. All of these cir-cuits are in lead cable and are mostly under-ground. No heat-coil protection is used, but standard 300-volt gaps are installed on both wires at all intermediate central offices and at terminals. The pilot circuits are continuously and automatically supervised.

In choosing a protective-relay scheme for a given application there are many factors to be considered, involving the financial,, engineering, and operating aspects. One of the latter factors which a relay engineer should not overlook in deciding between, for example, carrier pilot relays and pilot-wire relays, is the operating results reported for both types of protection. The figure of 96.5 per cent correct operation reported in this paper is good, but in this writer's opinion is not good enough. It does not equal the figures which have been reported frequently for other types of protection.

It is important to inquire further into the causes of the incorrect operations. It is unfortunate that the committee compiling this paper was unable to do this. No at-tempt was made to report all causes of incor-rect operations, and the only troubles given are causes of interruptions to the pilot-wire circuits. These may or may not have been

380 Discussions A I E E TRANSACTIONS

the causes of the wrong relay operations. Furthermore, no statement has been given as to the cause of the short circuits, open circuits, grounds, and so forth, on the pilot wires. This is very pertinent information and might have a decided bearing on one's choice of a system of relay protection, espe-cially for application to long transmission cir-cuits. This would be especially true if the causes of the pilot-wire trouble were found to be proportional to the physical lengths involved.

One of the causes which falls in the latter category in the writer's experience is the matter of pilot-wire troubles caused by what we classify as "interference by personnel." All of our pilot circuits are very plainly tagged in all central offices as being of the emergency type and not to be opened or tested without special permission. Yet several failures of the pilot circuits have re-sulted from neglect of such warnings at these locations. Furthermore, certain physical locations of pilot circuits, such as in man-hole splices, render special identification of pairs practically impossible. Specifically, a total of 14 failures of pilot-wire circuits from this cause has been experienced in a total of 748 circuit-mile years of operation.

It can be stated therefore that the use of facilities of another company in the main-tenance of devices which affect the opera-tion of a protective-relay scheme obviously increases the personnel errors which are be-yond the direct control of the operating company. This is one reason why fault-detector relays are installed sometimes on important installations to avoid incorrect operations caused by pilot-circuit failures.

This is not intended to disparage the co-operation of the engineers of the telephone company involved. On the contrary, their co-operation has been of the best, and con-ditions are definitely improving in our ex-perience. Nor is it the intention in this discussion to imply that our experience with leased pilot circuits, can be classed as unsatis-factory. In general the service has been very satisfactory, and in all fairness it should be stated that the difficulties described occurred mainly during the early periods of operation. It is intended only to point out that when utility operating engineers are deciding on protective schemes for important power-system units, the probability factor of such occurrences as described previously should be considered in the analysis.

E. L. Michelson (Commonwealth Edison Company, Chicago, 111.) : The survey which was made by the relay subcommittee in-cluded all outages of pilot wires, such as those caused by circuit rearrangements and replacement of cable in which the insulation has reached a dangerously law level. If the survey were limited to only those interrup-tions caused by an actual breakdown of the pilot wires while the main line was carrying load, the results would be much more favor-able for pilot-wire relaying. The experience in our company indicates that less than five per cent of the total number of interruptions are those caused by pilot-wire breakdowns. All the others are planned interruptions, which means that the main transmission line is deprived of the pilot wires only during those periods when the need for the main line is minimum. Therefore, in this respect, the continuity of pilot-wire circuits is much

better than that indicated in the report of the subcommittee.

W . K. Sonnemann: Mr. Brownlee's dis-cussion mentions the use of three-winding neutralizing transformers as a remedy for induction problems. It should be remem-bered that there is also available a scheme using a two-winding neutralizing transformer which is equivalent to a three-winding neutralizing transformer for differences in station ground potential and which handles longitudinal induction as well. As a matter of fact, the protection of pilot-wire circuits has recently been discussed in a paper "Protection of Pilot Wire Circuits" by E. L. Harder and M . A. Bostwick in A I E E TRANSACTIONS for September 1942. 1

There is a difference in opinion, ap-parently, as to how the relay used with the pilot wire should perform in the event of a faulty pilot wire. Some operators purposely set the relays below load currents for in-herent supervision of the pilot-wire circuit, so that if the pilot circuit is faulty, the cir-cuit breakers will trip.

It was pointed out that phase and ground-fault protection over only two pilot wires has been available by the directional comparison means for a longer period of time than the new two-wire a-c schemes. Per-haps the recent popularity of the two-wire a-c schemes is explained on the basis of their simplicity compared with most directional comparison schemes.

In a continuing investigation of this sub-ject, I feel that the causes of pilot-wire short circuits, open circuits, and grounds should be looked into as mentioned by Mr . Sleeper. Since this investigation primarily concerns the pilot wire itself as a unit in the protec-tive scheme, I do not believe an investiga-tion of faulty operations of the over-all scheme which may occur and which are attributed to defects or accidents not directly associated with the pilot wire should be considered by this working group.

REFERENCE

1 . P R O T E C T I O N O F P I L O T - W I R E C I R C U I T S , E . L. Harder , M . A . Bostwick. A I E E T R A N S A C T I O N S , volume 6 1 , 1 9 4 2 , September section, pages 6 4 5 - 5 2 .

A New Generator Differential Relay

Discussion a n d author's closure of paper 4 3 - 1 5 b y A . J . M c C o n n e l l , presented at t h e A I E E nat iona l technica l m e e t i n g , N e w Y o r k , Ε . Y . , January 2 5 - 2 9 , 1 9 4 3 , a n d p u b l i s h e d in A I E E T R A N S A C T I O N S , 1 9 4 3 , January s e c t i o n , pages 1 1 - 1 3 .

E. L. Harder (Westinghouse Electric and Manufacturing Company, East Pittsburgh, Pa.) : The paper by Mr. McConnell is commendable in presenting a brief yet com-prehensive introduction of a new member of the relay family circle. If applied within its proper application limits it should prove a useful member.

The paper could have been even shorter

and far more valuable had it omitted raising the question of a blind-spot danger with a variable slope relay. Every relay has its limit of application, this new relay included, and the description of what a relay may do at currents far beyond any at which it is ever applied is neither an asset nor a liability to the relay, and certainly no hazard to the application.

Variable slope means simply that as the current through the operating and restrain-ing windings is increased, the restraining torque increases in percentage. At small currents the restraint may be negligible, at larger currents it may amount to 10 per cent of the operating torque, at still larger currents 25 per cent, 50 per cent and so on. Obviously, if the current were carried on up to a sufficiently large value, and if the re-straining electromagnets did not undergo saturation in the meantime, a point might be reached where the restraining torque was in excess of the operating torque so that the relay would not operate. The whole ques-tion is where does this point occur, if it oc-curs at all?

One variable slope differential relay is designed for operation on secondary cur-rents, either symmetrical or asymmetrical, having an rms symmetrical component up to 100 amperes, whereas the crossing point oc-curs at or above 150 amperes. As the relay is applied with current transformers that result in not over 100 secondary amperes (rms symmetrical component) in order to secure uniformly high speed operation, there is absolutely no danger of misoperation from this source whatsoever. This relay was developed several years ago for bus pro-tection and to the best of my knowledge has a perfect record of operation both on staged tests and in the field.

W . K. Sonnemann (Westinghouse Electric and Manufacturing Company, Newark N. J.) : The author of this paper has solved an old problem in a somewhat different man-ner. The use of product restraint combined with the directional feature is interesting and does appear to have limited possibilities. However, it must not be concluded that this is the immediate and simple answer to all of the differential relay problems. It is still not possible to apply generator dif-ferential relays, particularly of the high-speed type, without due consideration being given to the current transformer character-

Figure 1 . Performance of mismatched current transformers w h e n c o m m o n b u r d e n is high

1 9 4 3 , VOLUME 6 2 Discussions 3 8 1