Graham Bright: Mr. Dreyfus has brought out certain conditions which may be met in a mine, but conditions at different mines vary so widely that it is impossible to lay down a set of rules to suit all conditions.

Mr. Dreyfus gave some interesting figures in regard to costs of operation at various load factors. He carries the load factor up to 75 per cent, which is far beyond anything that can be obtained at the average mine. He also bases some of his figures on the assumption of good management. In actual operation we seldom get good management in regard to the power supply, since those in charge wish to spend most of their time in mining and shipping coal, and the power plant is left largely to take care of itself until a breakdown occurs and then every person available comes in and works until the trouble is corrected.

I think Mr. Wood brought out very clearly the fact that even with the same cost or greater, it pays a mine operator to purchase central station power. Mr. Wood has been connected with mining work for a number of years and his opinion on this subject should carry great weight.

In regard to Mr. Gassman's remarks about there not being much in the " worry and care being removed, I think that no person will question the statement that the worry and care of a substation operating from synchronous converter or motorgenerator sets, is not to be compared in any way with the " worry and care" which exists with the average mining plant with boilers, engines and piping in the condition that they are usually found in. These plants as a rule are unable to carry overloads of any appreciable amounts, while a modern substation will take care of overloads up to 100 per cent with no trouble. Mr. Gassman mentions the fact that the central station power need be off the line only seven days in a year to wipe out the saving which has been shown. I believe that a large power company which has power off the line for seven days in one year would be considered very inefficient, and you will find that most of the larger power companies would consider this an extremely bad record.

As Mr. Sykes brought out, the figures given in my paper are not theoretical, but are really based on actual facts obtained from a great many mines which had been investigated during the last few months.

Mr. Müller has asked a question in regard to the power factor and voltage regulation of the power system. A fairly good power factor is generally obtained by taking the supply of direct current for the mines through synchronous converters or synchronous motor-generator sets. These machines can be set for a leading power factor and will tend to compensate for the induction motors which are used for driving the fans and pumps. The question of voltage regulation does not come up very frequently, since the mine operator as a rule is satisfied if he gets plenty of power at a fairly steady voltage. The lighting about a mine is of such a nature that close voltage regulation is not usually insisted upon.

of the American Institute of Electrical Engineers, Boston, Mass., May 15, 1912.

Copyright 1912. By A. I. E. E.

NOTES ON UNDERGROUND CONDUITS AND CABLES

BY C. T. MOSMAN

The consideration of an underground conduit and cable system involves many points of interest besides that of the physical construction of the duct line, manholes, etc. The point of prime importance and interest is the load or current which may be safely carried through a duct, and this is intimately related to nearly every feature of construction of the duct system and of the cables.

The heat generated in the cables results principally from I'R loss in the copper, but in the case of single-conductor cables carrying alternating currents may be considerably augmented by IR loss in lead sheaths, and slightly by dielectric hysteresis in the insulation.

The IR loss of large sizes of cable will be greater for alternating than for direct currents, due to the effective increase in resistance resulting from skin effect. Cables of 1,000,000 cir. mils or larger and for use on sixty-cycle service should be provided with a hemp core, to reduce the skin effect to a negligible value.

This heat will be dissipated by conduction through the duct structure to the surrounding earth; by conduction along the copper lead to the manholes, and there dissipated by the ventilation, if any; by air currents through the ducts from manhole to manhole. There will be a continuous fall of temperature through the various mediums from the copper to the earth, so that additions must be made to any exterior temperature in order to arrive at the temperature of the copper.

The advisable limit of temperature of the copper differs according to the type of insulation used and the voltage impressed

thereon. To avoid undue deterioration, rubber insulation should not be operated above 50 deg. cent. and varnished cambric or paper insulation above 80 deg. cent., and this latter figure should be understood to apply only to low-potential operation.

The insulation resistance of insulating material varies with the temperature, and differently for different insulations. The rate of decrease of insulation resistance, with increase of temperature, of high-quality rubber insulation is much lower than is the case with either paper or varnished cambric, the rate of decrease of paper being about four times that of good rubber, and the rate of varnished cambric about eight times.

The variation of resistance to puncture with variation of temperature is of much more importance than the variation of insulation resistance. The puncture resistance varies with the time of application of the potential. The instantaneous resistance to puncture of varnished cambric insulation is the same hot or cold, but above a critical temperature of 70 to 80 deg. cent. the sustained puncture resistance decreases with increase of temperature, while below 60 deg. cent. this effect is negligible. This effect seems to depend upon variations of dielectric hysteresis with temperature and it varies in percentage with the thickness of insulation.

This effect is not of practical moment in the case of cables built for operation at 2000 volts or less. With cable built for 2500 volts the effect may be approximately 5 per cent, with 10,000-volt cable about 20 per cent, and with 25,000-volt cable about 25 per cent.

It should be noted that the above-stated decreases of puncture resistance exist only while the high temperature exists, the insulation regaining its normal resistance when cooled.

Paper insulation is subject to the same effect, but to a somewhat lesser degree.

Rubber insulation maintains its normal value of puncture resistance much better than either paper or varnished cambric, but suffers a permanent deterioration.

Consideration of the above indicates that it would be conservative to limit the copper temperature of high-tension cable to 50 deg. cent. for any of the above types of insulation.

The thermal drop through the insulation will vary with the character and thickness of the insulation and the temperature of the copper, and the rate of dissipation of energy. Tests have been made on various types of insulation used on generator coils to determine the thermal drop. The results of tests by different experimenters differ widely.

One set of tests gives values of thermal drop in degrees centigrade per inch thickness of insulation, per watt per square inch dissipated, of 169 for varnished cambric tape at the rate of 0.30 watts per sq. in. (6.45 sq. cm.), and 136 at the rate of 1.63 watts per sq. in. (6.45 sq. cm.) From these figures it would appear that at rates of radiation common in cable practise the thermal drop would be in excess of 169 deg. cent. per watt per square inch per one in. (2.54 cm.) thickness.

Applying the above to a varnished cambric insulated cable with a rate of 0.05 watts per sq. in. (6.45 sq. cm.) and insulation of 11/32 in. (8.7 mm.) thickness, the copper may be expected to be at about 3 deg. cent. higher temperature than the surface of the cable, thus showing that this is not an important matter in connection with the average commercial cable. The drop in temperature through the air in a duct and along the length of the duct will depend upon the opportunity for a definite circulation of air through the duct, which will usually be negligible. The

elevation, the ducts next to earth at the bottom and sides being the coolest and those furthest removed from earth the warmest.

The foregoing must serve as my excuse for thinking that a reference to some actual tests made on a conduit system may prove of interest.

A manufacturing plant installed a conduit system in the mill yard, using fiber conduit laid in cement. Because of physical limitations the 81 ducts were laid nine wide and nine deep. Only a few cables were at first installed, but when later it became necessary to add to the number, it seemed wise to investigate somewhat before deciding upon the cable size and the permissible load per cable.

The main conduit line extends from manholes No. 1 to No. 9 as shown in Fig. 1.

Referring to Fig. 3, giving a section of the conduit line, it will be noted that all the cables installed are single-conductor

1,000,000-cir. mil, paper-insulated, except those in ducts 13, 14 and 15, which are 500,000-cir. mil single-conductor, rubber-insula ted.

Of the total length of main conduit line, the sections between manholes No. 1 and No. 2, and manholes No. 7 and No. 9, are the only sections which are actually underground. The balance, between manholes No. 2 and No. 7, are above the basement floor level in one of the buildings; the lower row of ducts being some ten to twelve in. (25 to 30 cm.) above the floor level.

The latter sections, being inside the mill basement, with three sides of the system exposed to air, may be expected to operate at higher temperatures than the former sections, which are

VANHOLF
NO. 3

MANHOLE

NO. 4

D

FIG. 2-THERMOMETER LOCATIONS, CONDUIT LINE, MANHOLES 3 to 4.

buried in moist earth. Further, these sections may be expected to have always a greater number of ducts occupied by loaded cables than sections 7 to 9. The limit of the capacity of the duct system hence should depend upon the hottest portion of these indoor sections. Thermometers were placed on the lead sheaths of the cables at the duct ends in manholes Nos. 7, 6, 5 and 4. Devoting a day to each manhole, readings were taken hourly of temperatures and loads of the cables.

The results of these tests indicated that the highest temperature would be obtained at manhole No. 4, and the section. from manholes No. 3 to No. 4 was finally taken as the one representing probably the severest condition as to temperature.