May 22, 1914. THE COLLIERY GUARDIAN. 1125 The Action of Acid Mine Water on the Insulation of Electric Conductors. From a preliminary report {Technical Paper 58) issued by the United States Bureau of Mines. By H. H. CLARK and L. C. ILSLEY. The purposes of this preliminary investigation were to determine the character of acid water to be used in future tests of this sort; to ascertain the most effective method of applying the water to the insulation in order to determine, for the insulated conductors to be tested, the minimum dimensions that would permit the making of satisfactory observations of the action of the acid water upon the insulation; to discover the best way to prepare and mount for test the insulated conductors to be investigated; to find out something of the acid water resisting qualities of several kinds of insulation; and, finally, to disclose as many as possible of those unforeseen obstacles that always appear during the pro- gress of a new investigation and cause annoying and expensive delays. The collection and analysis of water from a number of mines showed that it was not possible to obtain, even from the same mine, two samples of water that had the same constituents combined in the same proportions. The acidity of mine water changes continually and is greatly affected by the surface rainfall and by other factors beyond the control of the investigators. The only deleterious constituent of the waters, other than sulphuric acid, was iron sulphate. A water containing iron sulphate and sulphuric acid in any desired propor- tions could be easily prepared and kept in a satisfactorily uniform condition. It was therefore decided to use in the investigations artificial solutions of this sort, as there seemed to be no doubt that the use of such solu- tions would give more accurate and consistent results gallons) of ferrous sulphate. This solution is designated as “ solution No. 2.” Materials Tested. , Samples of four kinds of insulating material were tested as follows :—Special varnished cambric, 30 per cent. Para rubber, standard varnished cambric, and paper protected by a lead sheath. Special Varnished Cambric.—The special varnished cambric samples, termed “special samples” and designated by the letter S consisted of No. 6 Brown and Sharpe gauge solid wire, insulated with varnished cambric specially treated to withstand moisture, and protected with a single cotton braid saturated with compound, but not having the smooth, hard finish usually seen on the so-called weatherproof braids. The dimensions of these samples were as follows :—Length of each sample, 50 ft.; diameter of wire, 0*162 in.; thickness of cambric insula- tion (eight layers), 0’146 in.; thickness of single cotton braid, 0’056 in. Thirty Per Cent. Para Rubber.—The 30 per cent. Para rubber samples, termed “ rubber samples ” and designated by the letter R consisted of a No. 1/0 Brown and Sharpe gauge) solid wire insulated with 30 per cent. Para rubber compound in accordance with the specifica- tions of the National Rubber Covered Wire Engineers’ Association. The rubber insulation was protected with a single wrapping of tape and a double cotton braid having a weatherproof finish. The dimensions of these samples were as follows :—Length of each sample, 50 ft.; standard varnished cambric and treated with solution No. 1; 2S2 designates sample No. 2 insulated with special varnished cambric and treated with solution No. 2. The conductors to be tested were placed in two wooden tanks 13 ft. 8 in. long by 5 ft. 4 in. wide and 6 in. deep. Three trays were placed in each tank, each tray con- taining five samples of one kind of insulation. One of the trays in each tank also contained a sample of lead- sheathed paper-insulated telephone cable. One end of each tray rested on the bottom of the tank and the other end of the tray was raised about 8 in. Thus the floor of each tray was inclined so that the water would drain off. The trays were installed in such a manner that they could be lowered into water in the tanks whenever it was desired to take insulation readings or make high- potential tests. The samples in tank No. 1 were treated with solution No. 1 and the samples in tank No. 2 were treated with solution No. 2. Beginning January 7, 1911, the samples were treated daily with the proper solution, and once every month or six weeks readings were taken of the insulation resistance of each sample. The solutions were applied to the insulation of the conductors by sprinking from an ordinary watering pot. For the first four weeks the sprinkling was done three times daily, for the next 10 weeks the sprinkling was done once daily, for the next 76 weeks the sprinkling was done twice each day, and for the rest of the time the sprinkling was done once each day. The amount of solution applied at any one time was approximately 0’17 oz. for each linear foot of conductor. In addition to sprinkling, aspirator bottles were arranged so that the solutions could drip slowly upon certain con- ductors, thereby maintaining a saturated condition of the insulation at one point. One sample in each tray was also embedded in coal dust, in order to simulate a possible underground condition. Beginning eight months after the tests were first started, high-potential tests were made upon each cam- bric, rubber and special sample. A record was kept of the temperature of the room in which the testing tanks Table I.—Effect of Solution on Appearance of Exterior of Samples. Table II.—Decrease of the Insulation Resistance of the Samples as Treatment Progressed. Date. Exterior appearance of— Rubber samples. Cambric samples. Special samples. Tank 1. Tank 2. Tank 1. Tank 2. Tank 1. Tank 2. Unfilled braid. Jan. 7,1911 Feb. 10,1911 Mar. 13,1911 Apr. 20,1911 June 8,1911 July 21,1911 Sept. 9,1911 Oct. 27,1911 Dec. 18,1911 Jan. 29,1912 Mar. 15,1912 May July 20,1’312 ScAt 21,1912 Nov. 25,1912 Good weatherproof finish... Smooth, glossy, weather- proof finish. Threadbare....................do.................. Extremely threadbare............do................... Weatherproof covering has become very shabby Little change................ Cracks have Cracks have begun to begun to develop. develop, but not so many as in tank 1. Continuous cracks along the tops of the samples. No marked change........ ....do..................... Cracks have increased...... Cracks have increased and second covering has broken apart in many places. Both layers of weatherproof covering on the top of the sample practically de- stroyed. ____do...................... No change observed......... .....do..................... ____do...................... Surface of weatherproof covering appears duller than before. Surface of weatherproof covering more thread- bare than before. Surface of weatherproof covering has become more threadbare. No marked change........ ____do................... Weatherproof covering seems softened. Covering has become rot- ten, but its appearance has not materially changed. Cracks began to appear.... No change observed....... ....do.................... ....do.................... Braid eaten at bends; crack- ed along top of sample. Long cracks in braid on top of sample. Cracks have increased. Cracks in first layer of cam- bric. Braid eaten at * bends; looks ready to crack but is still intact. Few cracks; braid looks weak and crumbly. Braid decayed and can be brushed off. Practically no change. No change Do. Do. Do. Do Practically no Braid has be- change. come more decayed. No change observed. Do. Do. Do. Date of measure- ment, a Temper- ature of water. Insulation resistance in megohms. Rubber. Standard cambric. Special cambric. Lead-sheathed paper. & Tank 1. Tank 2. Tank 1. Tank 2. Tank 1. Tank 2. Tank 1. Tank 2. 1910. °C. ' Dec. 30 24 46,300 49,100 2,400 2,085 1,340 1,490 1911. Fob. 9 24 33,400 32,800 1,217 915 405 393 Mar. 11 c23 30,740 32,040 802 692 383 436 750,000 74,BOO Apr. 18 <3 24 26,940 29,000 446 387 282 316 1,215,000 63,COO J:izo 6 24 21,840 22,830 182 143 153 980,000 59,700 Jana 7 24 20,540 93.3 106.5 166 1,390,666 July 18 24 22,310 77.9 112 66,500 Sept. 7 24 21,660 23,510 58.1 47.5 89.9 88.1 545,000 74,400 Get 25 24 21,600 23,320 38.5 29.1 72.0 69.4 865, (XX) 48,300 Doc. 12 Dec. 14..... 24 24 19,190 20,440 20.3 17.1 78.5 58.5 690,000 70,000 1912. Jan. 26 «24 19,050 23,780 13.2 11.9 58.5 52.2 870,000 Mar. 13 24 20,300 21,000 9.3 8.5 63.1 46.7 864,000 77,300 May 28 24 21,380 19,660 /6.0 4.6 45.7 29.5 536,000 67,000 July 24 24 21,780 20,540 g 4.5 /3.5 36.2 24.1 752,200 75,300 70,600 Sept 17 24 19,680 18,960 03.0 02.7 25.5 18.0 1,386,000 Nov. 20 24 22,380 21,840 *3.0 <2.0 25.9 17.4 795,200 69,270 1913. Jan. 14 24 22,800 22,400 <3.6 <1.9 42.0 16.5 795,200 67,200 g Treatment with acid water was begun on January 7,1911. & Only one sample in each tank. This material was not received until the tests of the other samples had been started. The seeming discrepancy in the insulation resistance from month to month is explained by the fact that the resistances were usually too great to be measured with much accuracy without the exercise cl more care than the purposes of this particular part of the test demanded. The sample in tank 1 was see mingly more carefully prepared by the manufacturers than the sample in tank 2. This accounts for the difference in insulation resistance. c This figure refers only to tank 1. The temperature of tank 2 was 22.6. ^This figure refers only to standard and special cambric samples. Temperature of rubber samples in tank 1 was 23.75 and in tank 2,23.3. e This figure refers to tank 1. The temperature of tank 2 was 24.2. / Average of measurements made on 4 samples. ? Average of measurements made on 3 samples. K' Average of measurements made on 2 samples. $ Rscdhig taken on the only sample left. than could be obtained by the use of mine waters of varying qualities and compositions. It was decided to use two solutions, both to contain sulphuric acid in equal amounts, but only one to contain iron sulphate. It was decided that the most effective way of applying the acid solutions to the insulation under test would be to sprinkle the insulation with the solution at such intervals as would allow the solution time to evaporate between each application. The conductors were mounted so that no leakage paths could exist around the insulation or across its surface. It was decided to test the effectiveness of sealing both ends of each conductor in a manner later described. In order to simulate the conditions of actual practice, it was decided to make bends at several points in the conductors. It was decided to mount the conductors in a horizontal position in order to facilitate the application and absorption of the acid solutions. The average amount of free sulphuric acid found by analysing samples of 16 different mine waters was 492 parts per 1,000,060 (23’68 grains per gallon). The maximum amount of free sulphuric acid found in any sample was 3,662 parts per 1,000,000 (212’49 grains per gallon). . It was decided that the artificial waters should contain a little more acid than the strongest mine water, in order to hasten results and have a margin of safety in the matter of acidity. A solution was therefore pre- pared containing 4,500 parts per 1,000,000 (262’35 grains per gallon) of free sulphuric acid. This solution is designated as “solution No. 1.” This solution was prepared containing the same amount of acid and in addition 12,000 parts per 1,000,000 (699’6 grains per diameter of wire, 0’325 in.; thickness of rubber insula- tion, 0’0925 in.; thickness of tape outside of rubber, 0’009 in.; thickness of inside braid, 0’035 in; thickness of outside braid,.O’033 in. Standard Varnished Cambric.—The standard var- nished cambric samples, termed “ cambric samples ” and designated by the letter C consisted of No. 6 (Brown and Sharpe gauge) solid wire, insulated with varnished cambric and protected with a double cotton braid having a weatherproof finish. The dimensions of these samples were as follows :—Length of each sample, 50 ft.; diameter of wire, 0’162 in.; thickness of varnished cam- bric (11 layers), 0’116 in; thickness of tape next, to cambric, 0’006 in.; thickness of inside braid, 0’029 in.; thickness of outside braid, 0’04 in. Lead-Sheathed Paper.—The lead-sheathed paper samples, designated by the letter T were prepared from a 5-pair, lead-sheathed, paper-insulated telephone cable. The dimensions of these samples were as follows :— Length of each sample, 25 ft.; thickness of lead sheath, 3-32 in.; size of wire, No. 22 Brown and Sharpe gauge. Two layers of paper were wrapped around each con- ductor, and the core formed from the 10 conductors was also wrapped with manila paper. This telephone cable was tested to ascertain the action of the acid upon the lead sheath. Paper insulation was selected because it would show a high insulation resistance, while the sheath was intact, but would show a very low insulation resistance as soon as the slightest hole was made in the lead sheath. Symbols were adopted to designate the samples. For example, 1C4 designates sample No, 4 insulated with were installed, and during the insulation resistance readings the temperature of the water was maintained constant. Testing Equipment. The outfit for measuring the resistance of the insula- tion consisted of a battery and a high-sensibility gal- vanometer used in conjunction with a lamp and scale, a one-tenth megohm resistance box, an Ayrton shunt, a reversing switch, and a contact key, the four last named being mounted on a hard rubber base. The probable error in determining the insulation resistance of the rubber insulated samples was about 3 per cent. The probable error in determining the insulation resistance of the other samples was less than 3 per cent. It is believed that the errors due to temperature and current leakage are negligible. The high potential testing equipment consisted of a 5 kw., 30,000-volt testing transformer with an especially calibrated potential transformer for reading the voltage across the high tension terminals. The voltage variation ■wf s obtained by the use of an induction regulator con- nected in series with the low-tension side of the testing transformer. Results of Tests. The table above gives the results obtained from the periodic measurements of insulation resistance. The values given are the average of the samples under test in each tray. There were five samples in every instance unless the contrary is stated. Exterior Appearance of Samples.—The samples were examined after the completion of each periodical test,