ficient conductance in the negative return, improper use of such conductance, excessive feeding distances and other causes of electrolysis trouble can usually be definitely determined.

B. TYPES OF SURVEYS

In the following discussion several types of surveys must be recognized. The first is that which may be called a complete electrolysis survey which is made for the purpose of determining the extent and location of the danger areas, and with a view of determining the proper procedure to be followed for the mitigation of any trouble which may be found to exist.

The second type of survey, known as a maintenance survey, embraces such surveys as would usually be made by a pipe or cable owning company solely for the purpose of determining whether previously existing conditions have changed, and differs from the more complete survey mainly in that most of the information with respect to the railway power distribution system is not required and fewer electrical measurements are taken, the number and character of such measurements depending on the thoroughness with which the survey is to be carried out.

A third type of survey which needs little discussion, except as to methods of making tests, is one made to determine whether ordinances or regulations governing electrolysis conditions in a municipality are being complied with. Such surveys are usually made periodically, in periods varying from three months to a year. In general, only those quantities are measured which are specifically defined by the ordinances or regulations which are in effect in the locality in question.

C. GENERAL PRELIMINARY DATA

1. Data on Underground Structures.

In making electrolysis surveys, a considerable amount of preliminary data are usually desirable. It is important first to gather all evidence regarding the character, extent and location of known damage to underground structures. This evidence is usually obtained from the utility companies concerned, but even though these companies can give no direct testimony as to the injury to underground structures, this should not be taken to indicate that no damage exists. The data on the underground systems should include the relative location of the mains, the railway tracks and underground cable systems. The size and kinds of pipe and the types of pipe joints used are usually important. Numerous questions relating to the interconnection of gas, water, and cable systems are also of importance.

2. Data on Railway Systems.

As regards the data on the railway systems, the following should be determined: (1) Location and capacities of direct current railway supply stations; (2) Location of railway lines and character of service, whether city, suburban, or interurban, and the car schedules on different parts of

the system. This latter will have a bearing on the length of time necessary for taking readings at various points in order to get representative results; (3) Physical data on railway tracks, such as size of rails, types of bonds and joints, and character of roadbed construction; (4) Practice of the railway company in regard to crossbonding, bond maintenance and bond testing; and (5) Miscellaneous data. In most cases it is desirable to have all-day load curves to facilitate the interpretation of data taken over short intervals at various hours of the day. Where the load varies considerably in different sections of a power house feeding area, it may be necessary to get the load curve on different feeders in some cases. Where a survey is made with the ultimate purpose of correcting electrolysis conditions by applying some method of mitigation, it will be necessary to secure complete data on the magnitude and distribution of the load, the substation and feeder systems, frequency of schedules and probable future growth of traffic.

D. COOPERATION IN MAKING SURVEYS

Special surveys for determining whether ordinances are being complied with and maintenance surveys can usually be made by any particular utility interested. Complete surveys, however, which are to be preliminary to the application of electrolysis mitigative measures should preferably be carried out on a cooperative basis by the various utilities interested, including both the railways and owners of underground utilities. It is of the utmost importance that a comprehensive plan of procedure be followed, so that all information relating to the electrolysis conditions of all of the underground structures may be available in the planning and carrying out of the test. In order to bring about such a unification of data and methods, it is necessary to have the full cooperation of all utilities whose properties are affected by electrolytic conditions. In general, this cooperation can best be brought about by having the electrolysis survey and the mitigative measures, if any are to be applied, designed and installed under the jurisdiction of a joint committee representing all of the interests concerned or at the discretion of such committee by an engineer employed by the committee, or jointly by the parties to the survey.

II. Electrical Measurements

The electrical measurements to be made during an electrolysis survey may be logically classified in either of two ways, namely, (1) on the basis of the structures on which the measurements are to be made, that is, whether on the railway system, pipe system, or cable system, and (2) on the basis of the character of the measurements, whether of voltage, current on, or current leaving a structure, etc. Inasmuch as several or all of the various types of electrical measurements may at times have to be made on all of the affected structures, and since the methods used will be substantially the same regardless of the utility system to which they apply, it appears most logical to follow the latter classification and discuss the subject from the standpoint of the character of the measurements to be made.

A. VOLTAGE SURVEYS

The number and character of the potential readings required depend on the information desired. As previously pointed out, the readings depend on the thoroughness of the investigation to be made, and it is to be understood that many of the measurements described below would. often not be necessary, and in general would be taken only during the course of a complete electrolysis survey. Voltage surveys are here divided into two main classes: (a) Voltage measurements between two points on the same structure, and (b) Measurement of the potential difference between structures.

1. Measurement of Maximum Potential Drop Along Railway Structures.

(a) Importance of Maximum Potential Drop Measurements. Such measurements, when interpreted in the light of other conditions to be discussed later, afford a valuable index to electrolysis conditions generally. It is, further, one of the easiest quantities to determine in an electrolysis survey if the use of telephone lines can be secured. These measurements show in general whether the railway system is properly maintained and what lines or sections are most in need of repair and rebonding. When taken in conjunction with the load data, they may be used for an approximate calculation of power losses in the railway return and when studied with due regard to the character and location of railway lines and supply stations, together with the distribution of load, they afford a valuable index as to the need of, or the modification of the track feeder system. It is therefore desirable, as a rule, to take a good many of these measurements as a part of any complete electrolysis survey.

(b) Procedure in Making Maximum Potential Drop Measurements. The first step in making measurements of this kind in a city is to determine the location of points between which potentials are to be observed. These usually comprise points on the track most remote from the power supply station as the points of highest potential, and the points on the track nearest the power station as the point of lowest potential. In some cases, however, especially where insulated track feeders are used, the point of lowest potential may be at the point of connection of one of the insulated feeders which may be at a considerable distance from the power station. It is desirable, as a rule, to measure the difference of potential between the points of connection to the tracks of all of the insulated track feeders in order that the points of lowest potential may be determined. It is desirable, as a rule, also to select points intermediate between the points of highest and lowest potential so that the distribution of the potential drop may be determined, as this will give a valuable insight into the location of bad stretches of track and of concentration of return current in the tracks.

Reference should be made to a positive feeder map from which a list of power stations and their approximate feeding distances can be determined. Special lines to these points may be run or spare wires may be borrowed or leased from the telephone company. In the latter case,

the continuous cooperation of the telephone company is required. Having a list of points to be reached, the telephone representatives can prepare a table showing the terminal boxes and numbers of spare pairs, including trunk lines which are necessary to make a complete circuit between one of the telephone central offices, or other suitable central point where the measuring instruments are to be placed and the points to which measurements are to be made. All measurements can then be made between the point of lowest potential and all other points selected, and between any two points as desired. Temporary circuits are necessary when no spare telephone conductors are available, in which case working conductors may sometimes be used for short periods. In most cases, it is desirable to make permanent connections to the track for maximum voltage drop measurements, but where, for any reason, permanent or semi-permanent connections to the track cannot be made, temporary connections become necessary and the installation of these temporary connections will require considerable time and expense for labor.

It will be found most convenient to bring all the lines from the various points on the track network to a large board on which is mounted a map of the railway system, each wire being fastened to a binding post located at a point on the map corresponding to the point on the track from which the wire comes. Once the correct connection of wires has been verified, one can readily connect the voltmeter to wires leading to any points in the city without possibility of error.

While making track voltage measurements, and, in fact, all other measurements, it is desirable to arrange to have the test data worked up and tabulated so that it can be carefully studied as the work progresses. This is important because, as pointed out above, the tests to be made during the course of the survey often depend in large measure on the results of preliminary measurements so that by making a study of the preliminary data while the work is in progress, it is often possible to modify original plans in such a way as to greatly increase the value of the test data.

In making electrolysis measurements, it is desirable to take readings at each point over as long a period as practicable. Owing to the great variability of railway loads, it is important to have the readings cover at least one complete cycle of the load, and often several complete cycles are desirable. It is quite common practice to take such readings over a period of one hour, but in some cases, especially on interurban lines and others where the schedule is very infrequent, still longer periods may be necessary. Even where readings are taken over one hour it will generally be necessary for comparative purposes to reduce these readings to an equivalent twenty-four-hour value, and in some cases also corrections have to be made for seasonal changes of the load. This matter will be treated at some length under the discussion of the interpretation of electrolysis survey data.

2. Potential Gradient Measurements.

(a) Scope of Term. Under the head of potential gradients will be included all potential measurements between different points on the track

or between different points in the earth spaced materially less than the extreme feeding distances within the powerhouse areas.

(b) Measurement of Potential Gradients in Tracks. Potential gradient measurements are usually made on the railway tracks, but at times also on pipe systems or even in the earth. The procedure will vary considerably because of the variability of the distances over which measurements are to be made. If the spans are long, telephone wires are the most convenient and the measurements are made in the same way as track voltage measurements described above. Where the distances are relatively short, however, as for example, a few hundred feet or less, a temporary wire between the two points of measurement will usually be. most convenient. Connections to the system on which measurements are made will depend on whether the tests are being made between points on the tracks or on the pipe system or between points in the earth. For measurements between points on the tracks or on the pipe system or other metallic structure metallic terminals may be firmly held against the rail or pipe or a wire may be swedged in a slot sawed in the pipe or rail under test. It is sometimes desirable to make potential measurements directly between two points in the earth, though the most common practice has been to take them on the track network. Special situations may arise where potential measurements between various points in the earth are more valuable than those taken on the underground structures. For example, the presence and direction of large transverse currents in the vicinity of important mains can be determined. Buried pipe lines or other conductors at uncertain locations which are discharging current into the earth may be located approximately by earth gradient measurements, there being a reversal or abrupt change in the gradient when the conductor is crossed.

In making earth gradient measurements between points relatively close together, it is essential that a pair of non-polarizable electrodes be used if a high degree of accuracy is to be attained. Such electrodes are now in process of development at the Bureau of Standards.

The periods over which gradient measurements should be made and the procedure in working up the data during the progress of the survey are governed by the same considerations as discussed above in the treatment of the track voltage measurements.

3. Measurement of Potential Differences.

(a) Purpose of Measurement of Potential Differences. Measurement of potential differences are beyond question the measurements most frequently made in connection with electrolysis tests and when their limitations are properly taken into account, they afford a valuable index to electrolysis conditions. It should be emphasized, however, that they are chiefly of qualitative significance, being valuable for indicating the region in which more or less damage to pipes may be in progress, but not giving any definite information as to the rate at which injury to pipes may be progressing. This is due to the fact that the resistivity of the earth and railway roadbed varies with local conditions, that is, a given potential difference that would be practically safe under some conditions of soil