The problem of electrolysis is one of the most elusive and, therefore, most interesting of the many problems that the electrical engineer has to deal with. Its far reaching effects and attending dangers make it also of importance. In large cities especially, it is the cause of much loss, not only direct, through damage to pipes, cables and structures, but, also, indirect, through its weakening of water and gas pipes, causing loss of property or even life. If, during a large fire, a weakened pipe gave way, tileproperty lost through lack of water might be very considerable, or, should a gas-pipe be weakened or punctured, the leaking gas might easily cause loss of life through suffocation or explosion. For these reasons, if for no other, every precaution should be taken to prevent, or at least reduce electrolysis to the greatest extent. The popular conception of electrolysis is the destruction of metallic structures by the electric current flowing through them. This is incorrect. The actual flowing of the current through the structure does no damage so long as the quantity of it is not large enough to heat it. It is only the current leaving the structure that causes damage. In leaving, the current takes with it a portion of the metal. The amount removed depends on the kind 01 metal of which the structure is composed; in iron, for example, one ampere will remove approximately 20 lb. per year. For lead, brass, copper, etc., the amount removed varies considerably. One can readily see that should any amount of current be taken locally from any iron structure, a very considerable damage would occur in a very short time. Fortunately, however, the greater part of the current does not leave locally, although it does in most cases leave in well-defined areas. It is to this fact that the difficulty of prevention may be ascribed. It roust be stated, however, as reSfards cast-iron pipe that the current does have a tendency to localise more or less, due, no doubt, to the lack of homogeneity of the iron. This gives rise to the pitting which is so freijiiently noted on iron, gas and water pipes. There may be also local conditions which wili at times cause local action on pipes and structures of more homogeneous character. The major portion of the electrolysis is caused by the action of the return current from railway systems, though damage of like character has often been traced to grounded lighting systems. The current returning from the cars to the power station takes the easiest possible path. This may be direct to the station byway of the rails and grounded returns, or through the earth, or by means of a combination of all the conducting structures between the car and the power-house. Electricity fnllows well-defined laws as regards its flow, one of which, and that which is most important with reference to electrolysis, is that the current will divide itself among the paths open to it in inverse proportion to the resistance of these paths. This is the prime cause of all the difficulty, for the current, on reaching the rails divides, a portion of it remaining on these rails, the rest spreading and dividing between the earth and other structures in the vicinity. Thus, unless the rails form the shortest and lowest resistance path from the car to the power-house, the greater part of the current will be distributed among other structures. It will follow these structures until it reaches other structures giving a more direct and a lower resistance path to the power-house, or possibly. through some part of the return system of the railway. Where the current leaves, destruction is bound to occur. Where it flows along the water and gas-pipes, there is possibility of widespread damage all along its path from the very end of the trolley line to the power station, since the resistance of the pipe joints is very much greater than the resistance of the pipe itself, although the resistance of the entire pipe-line may be low enough to cause a very considerable flow of current over it. The effect of these high-resistance joints is to cause a part of the current to jump them, flowing from one side of the joint through the earth to the pipe on the other side. This leaving causes damage to each and every joint on the pipe line and will in time result in the destruction or puncturing of the pipe at the joints. A consideration of the above facts will make plain the necessity for making the return system of the railway as low in resistance as possible, since by this means the greater part of the current may be confined to the rails or returns. It is plain that attention should be paid to the placing of return conductors direct to the power-station along the natural line of flow, so that the return current may find at least as easy a path over the return as it would over any pipe or metallic structure in the vicinity from which it is advisable to keep it. These returns should be properly and carefully fastened to the rails and kept in good condition. It is evident that the track and return should be kept in good condition and of low resistance, since it is their province to take care of the return current from the railw-ay, and, since the railway is the cause of the trouble, it is the place of the railway to take every and all possible steps to prevent electrolysis. Rail bonds should be carefully applied, frequently examined and the joints kept in the very best possible electrical condition. The ground taps should be inspected frequently, and the entire return system kept in as good condition as possible. There is no reason why the railway should neglect this; in fact, there is every reason why they should not neglect it. A poor return system is quite as much a cause of lost power as a poor distributing system, and, second, they inay also be the sufferers to a considerable extent should a fire occur in one of their plants and a weakened pipe give out during the fire. Furthermore, water companies and gas companies, telephone and electric light companies, should not be expected to handle on their pipes and cables any more of the return current than that which is absolutely unavoidable. It is not to be expected that no current will leave the railway returns; but it is to be expected that the railway will keep these returns of such low resistance that by far the greater part of the current will return on them, and that at no time will the current flowing on any other structure be dangerous in its extent. In Germany, France and England, the railway companies are compelled by the law to keep the drop over their ground returns down to a certain maximum figure. If, in this country, the drop per mile of track maximum were limited to the same amount that there covers the entire length of track, a great deal of electrolysis would be prevented. It is advisable that the maximum drop per mile of track in this country should be limited, first, to a figure of approximately 5 volts and then reducing this as conditions grow easier. This is one of the first and most important steps for the doing away of electrolysis. When this point has been reached it will be far easier to take steps to reduce electrolytic damage and possibly obviate it very considerably. So long as the present return system is allowed, electrolysis will occur. In order to determine whether electrolysis exists or not, it is first advisable to map out the area over which the water or gas-pipes, or structure under investigation. is positive to any other structure in the vicinity. The maps should be made separately for each two structures between which the potential is taken on the same scale, on thin material in order that they may be superimposed for examination and comparison. It is not enough to take the drop of potential between the structure under examination and the return circuit of the railway, as it often occurs that the potential between the rails and this structure is such as to indicate that there is no possibility of damage while the potential between the structure and some other conductor in the near vicinity may be such as will indicate the possibility of damage. The accompanying map shows the general form of one of these potential surveys. This map is simply given to show the method of laying out the potential between the tracks and the water pipes, so that the area which is dangerous from direct electrolysis, or from current leaving the pipes, may be seen at a glance. That section outside of the shaded area may, however, be a dangerous zone from the current flowing in the pipe. It is also advisable to take readings along the entire length of the structure to determine if there is a current flow’ in the structure. If this is a continuous metallic structure, like a cable sheath, the flow of current will do no harm unless excessive. Under these conditions it is advisable to take simultaneous readings in a number of places along the cablo, in order to determine whether there is any of this current leaving between the locations where the readings are being taken, this flow being due to some other cable or structure in the near vicinity being of lower potential than the cable under examination. Any decrease in the current from one point to another will show conclusively that current must be leaving somewhere between the points of readings. This current may not leave locally, but may he spread over some considerable area. If the readings are taken on a water or gas-pipe, the same intimation will be made. There is, however, a possibility in this case which does not exist in the case of a solid conductor, and that is the damage to these pipes at the joints. In taking reading on water and gas-pipes, it is necessary to use considerable precaution to assure the readings being taken over an unbroken section of the pipe in order to obviate the high resistance either of the break or possibly of a joint in the pipe. The accompanying sketch shows how the connection should he made. Care should also he taken to take the simultaneous readings over a number of pipe sections, since the weight varies more or less per length of pipe. If it is possible to determine the w’eight of the section under examination, this should he done and the readings taken based on this weight. A table is appended showing the weight of standard water pipes. It must not he assumed that a high potential difference betw’een the rails and structure under examination, or between any two structures means that electrolysis is taking place, or that any great current flow exists. In fact, it may be assumed under most conditions that a high potential difference means small current flow, unless the ground is a comparatively good conductor and conditions decidedly out of the ordinary exist. It is usual to find a very considerable potential difference between the rails and other structures where the soil is dry and sandy 011 account of the high resistance imposed by this character of soil and the current flow, under these conditions, is usually negligible. It is also unwise to assume that a low potential difference means no electrolysis, for under certain conditions of soil, a low potential difference will produce a considerable flow of current and cause damage. The potential survey, therefore, simply indicates a possibility of the flow of current which may or may not be causing damage. These potential surveys, taken in connection with the condition of the ground can be used by one having experience in such matters, to good advantage, and serve to map out” the locality which should be subject to rigid examination. It is unwise, however, to jump to any conclusions from tests of this character alone. A very good indication of the possibility of current flow between the rails and the structure or between any two structures may be found by taking the potential drop between the structures with a moderately high-resistance voltmeter, and then again taking the drop with an ammeter inserted between the two structures, taking both voltmeter and ammeter readings under these conditions. The connecting link between the two structures through the ammeter should be of comparatively low resistance. Should the voltage be high without the ammeter in circuit, and low with the ammeter in circuit and a comparatively small current Dow, it may be taken as an indication that the current flow between the structures under normal conditions is small. If, however, there is a considerable rush of current and the voltage does not drop very materially on the inserting of the ammeter, it may be safe to assume that there is considerable flow of current between the structures at or near the point of application of the ammeter; and it is, therefore, advisable to take current flow measurements along the structure on either side of this point, in order to determine as near as possible the location from which the current flows. It may be, however, that the current flow is distributed over some considerable area in case the two structures run parallel for some distance in ground of good conducting capacity. Under these conditions. a prevention at this point is practically impossible; and means should be taken to limit the current flow over the structure and also to prevent the current flowing to the structure. This will mean an examination of the conditions existing at the point or points at which the structure under examination is negative to the tracks or other structures, from which it is possible a current might flow; and if possible. the prevention at this point, of the current leaving for the structure under examination. There are numerous Other tests which may he made to gather indications regarding electrolysis and the flow of current along the pipes, as well as the flow of current from one pipe to another. One of the tests for current flow along the pipes is a repetition of the test just indicated for flow of current between structures. This consists of taking measurements along the pipe-line by means of a voltmeter, getting the voltage between points, then inserting a low resistance ammeter in the circuit, and taking the potential under these conditions. If we assume that V represents the voltage without the ammeter; V represents the voltage with the ammeter in circuit: A represents the current on the ammeter: then A represents the amperes to be determined. A: A’ : :V : (V — V). This formula is not absolutely correct, but serves to indicate approximately the current flowing over the pipe. Care must be taken in making this measurement, that the plugs or contacts are made on the same line of pipe, since only under these conditions will any results be obtained. The same general rule holds here as regards the amount of current flowing on the pipe being indicated by variation in voltage, as will readily be seen from the formula. It will not be safe to assume, however, in the use of this method to determine ground flow, that this formula applies: for the inserting of low-resistance bonds changes conditions to such a marked degree that the current indicated by the formula is not even approximate. It only serves to indicate what may be going on. Abroad, there have been a number of very interesting experiments and pieces of apparatus devised for the purpose of measuring small differences of potential and current flows in the earth. These instruments are described in an article on electrolysis in the February number of “The Electrical Age,” and arc quite well worthy of examination. This article is by Albert F. Ganz. Mr. Ganz also calls attention to the unreliability of many of the reports made on electrolysis—these reports being based on potential differences only. There is but one absolute method of preventing electrolysis from railway return currents, and that is the absolute prevention of the current from leaving the cars, or in other words, a complete metallic insulated system. By “insulated,” 1 mean from the ground at all points. A number of cities have compelled the use of this system; hut the vast majority still allow the use of the ground return. So long as this continues, electrolysis will exist to a greater or less extent. One of the first steps to reduce electrolysis is to compel the railway companies to limit the ground drop to some definite value, and to fine them heavily for every time that a test shows this ground drop to be exceeded. These tests should be made at the expense of the railway company and made frequently. One method of preventing electrolysis by means of preventing the flow of current over the structure, is the use of insulating joints—these joints heing composed of wood, rubber or fibre, as the ease may be, inserted cither at the joints of the pipe or in straight sections of the pipe, the pipe having been cut in order to enable them to be inserted. Under certain conditions these joints serve their purpose, as, for instance, in dry sandy soil, where the resistance of the soil is sufficient to prevent the current flow about the joint. Under these conditions, one or two joints will serve to prevent the flow of any considerable amount of current over the pipe. To insert one or two of these joints, however, in a pipe line flowing through a marshy district, the soil of which may be either acid or alkali, will result in disaster, for the ground itself is a very good conductor, and the current will jump these joints, tending to make the damage more severe locally at the joints themselves. It is only by inserting a very large number of these joints in a pipe line through a marshy district, that any results will be obtained. The theory of the joint is the inserting of sufficient resistance in the pipe line to cut the current to a negligible quantity. The bonding of the pipes to the return at or near the station, is not to be recommended, although under certain conditions it seems the only possible thing to do, for so many of the railways absolutely neglect their ground return that it is impossible for the water company or the gas company to prevent electrolysis of a severe character in any other way. It may be, however, that the use of a high resistance bond will serve the purpose, and at the same time reduce the current flow over the pipe materially. These high-resistance bonds may be inserted at a number of points in order to dissipate the current through a wider area, and to prevent so large an amount of current flowing over the pipe line as would occur should but one bond be used. High-resistance bonds may be made very cheaply, and in many cases serve their purpose to good advantage. It is advisable, however, in inserting bonds, as well as in the use of insulating joints, to use great care; and, furthermore, that they should be inserted under the direction of one familiar with their use and with the general conditions existing throughout the system. High-resistance bonds are not, as might be supposed from the word “bonds” actual physical connections between the structures, but are a device used to make the current leave the structure through a wire transferring the damage to a plate to which the wire is connected, which is located nearer to the structure to which the current is tending to flow than is the structure being connected. The sole purpose of these bonds is to localise the current flowing in the structure to be protected and compel it to leave this structure witht ut removing any metal. The use of the high-resistance is to keep the current on the structure to be protected, as low as possible. It is interesting to note that many cases of electrolysis have been found to exist not due to the return current from railways, but from grounded lighting systems. One case in particular that came under my observation was where a water main had been punctured and the railway system called to account. Tests were made in the vicinity of the pipe and all showed the water pipe negative to the railway. Measurements of current flow on the pipe showed a considerable volume. The pipe was replaced and the second break came within a few days. Another examination was made by the railway with a result similar to the first, and they disclaimed any liability regarding the damage. The engineers, however, continued their investigation and discovered that the flow of current was in a direction opposite to that which might be expected, from the railway, and following the matter up discovered that the current came from a house some distance up the street. Following this clue, they studied the lighting system and discovered a ground existing in one tiouse and another ground in the earth near the location of the fault, which was in the conduit of the lighting company, just inside the house outside of which the damage occurred. Another amusing case called electrolysis, occurred on the line of a small railway in one of the New England States, where an engineer was called in to report on the possibility of electrolysis, in order to forestall a lawsuit being promised by a water company who claimed their pipes were being destroyed. The accompanying map shows the location of the tracks, power plant and the water pipes under consideration, and also shows the locations where the damage was claimed to have occurred. A careful examination of the ground was made and potential tests made to see whether there was any appreciable flow of current, At the locations marked it was impossible to detect any flow of current, even with a millivoltmeter; and there was no flow of current over the pipes. Over the total length of the main pipe line the current flow did not exceed 5 amperes as a maximum, and which for only a few seconds during the day, the average flow being negligible. The fittings at which the damage occurred are shown in the illustrations. The damage is unquestionably due to what is known as sand wash and doubtless came from the improper connections leaving slight leaks, which under considerable pressure allowed the water to come through. This water in sandy soil soon wore away the pipe, allowing leaks to occur. In connection with this case the engineer was given a very excellent opportunity to see how one unfamiliar with electrolysis and yet having some knowledge of electricity would explain the trouble occurring, and which was really the cause of the promised lawsuit. One of the men interested in the water company was an electro-plater by profession and, meeting the engineer at dinner, offered the following explanation for Ihe trouble: “This damage is undoubtedly due to tlu action of electrolysis. You can plainly see on these fittings the action of the electric current, similar to what takes place in a plating bath You doubtless know that underlying ibis whole town is a large pond of pure water; and you must also lie aware from your electrical training that pure water is a very excellent conductor of electricity. My theory is that the current, leaving the rails. Hows first to the water pipes, as they are nearer the subterranean pond ; flows along these water pipes until it reaches an easy path to this conducting body of water, and from this body of water, farther down along the line to the rails, and thence to the power station; the damage occurring where the current leaves tlie water pipes.” Tile above theory is quite untenable, owing to the fact that water, especially pure water, is a very poor conductor of electricity, and this water was to all intents and purposes pure. Having explained his theory, he asked the engineer for his, upon which the engineer replied he was glad to know the other side; hut until he had made a more careful examination and was able to report what he found, he should prefer to keep his theories to himself; that the water company would doubtless receive a copy ot the report front the railway when it was made. The engineer reported the damage due to sand wash, and nothing further has been heard with reference to a suit. 1 think it is fair to assume in case of water pipes that a very considerable portion of the so-called electrolytic damage is nothing more than sand wash. There are also cases of local action due to the use of iron, brass and copper pipes in the near vicinity in salty or acidulated ground; and it is advisable in making tests to very carefully examine the ground, pipes, etc., for their general conditions. 1 think, front the above, one can readily see that it is inadvisable to judge too much from potential readings alone; that the use of insulating joints is not always accompanied hv good results: that highresistance bonds and insulating joints should be installed by one familiar with their use, and no pains should he spared to make the railways put their tracks and returns in the very best possible condition and keep them so.


Owensboro. Ky.. having sold its water bonds for $25,000, will at once begin to extend its distribution System.




Report of Special Committee of the American Waterworks Association to the Detroit Convention, June, 1904.

Your special committee on Electrolysis respect fully begs to submit the following as its report: At the twentieth annual meeting of this association, held in Richmond in 1900, a special committee on electrolysis was appointed to prepare and submit a report setting forth the result of experience and facts regarding the action of escaping or released electric currents upon water mains and service pipes, and to include in said report recommendations for the guidance of the association in dealing with this problem, as well as to formulate, for its approval, an expression of the attitude of this association on this question. In order that every member of the association might have ample time and opportunity to study the report before voting upon it at the next annual meeting, it was published in full in the proceedings for tgoo. At the twenty-first annual meeting, held in New York in 1901, this report was accepted, and the following preamble and resolutions contained therein were, without a dissenting vote, adopted as an expression of the attitude of the American Water Works association on this question : “Whereas, I he report of the special committee on electrolysis shows that, wherever the single trolley system is in use, a portion of the return electric current will flow on the water pipes, and that, whenever water pipes become positive to rails or gas pipe or other underground metal structures, the water pipes are injured; and that, whenever water mains, whether positive or negative to the rails, convey electric current, they are liable to injury thereby, near the joints; and,

“Whereas, It further appears from the said report that there is no known practical method by which owners of water pipes can protect themselves against electrolytic injury resulting from the operation of single trolley cars, but that electric railways may, by means of the conduit system or of the double overhead trolley system, be operated successfully and economically and without any injury to the water pipes; now, therefore, be it.

“Resolved, That the American Water Works association, as a national organisation, herewith unanimously maintains that street railways have no right so to operate their cars as to cause injury and destruction to water pipes; and be it further

Resolved. I hat street railways now operating under the single trolley system ought to be required to remove their return current from the ground, and that, if they continue to operate by current transmitted from the power station or motors on the cars, they should be required to provide a complete metallic circuit of sufficient capacity to convey all the current, and in a manner absolutely to insulate it from tlie rails and from the earth.”

At the present time, electrolysis suits are either being tried, or decisions are pending in St. Paul, Dayton, Peoria, and Indianapolis; and the owners of water and gas pipes in a number of other cities are awaiting the results of the above cases before beginning similar suits. In addition to the double trolley roads mentioned in the report of 1900, others have, under the direction of the engineer officers of the United States army, been installed, and are in successful operation in Havana, Cuba, and vicinity; and the double trolley is being successfully operated in a number of European cities. The experience of the members of your committee, and of those interested in the preservation of underground pipes all over the country, has strengthened their conviction as to the correctness of the statements contained in their special report formulated in 1900, and as to the wisdom and propriety of the action of the American Water Works association in adopting these resolutions in 1901 as the official expression of its attitude on this important cpiestion. Your committee can, therefore, suggest no changes in its earlier report, while, on the other hand, it might submit an array of facts, the results of investigations and experience, which would fully substantiate the statements therein contained.




Special Committee on Electrolysis.