278 BELL SYSTEM TECHNICAL JOURNAL 



season along a 50-mile section of the Stevens Point-Minneapolis route. 

 The oscillographs, which were arranged to trip where the voltage exceeded 

 100 volts, recorded lightning voltages due to some 600 strokes on 38 days, 

 the Weather Bureau average being ii thunderstorm days for this region in 

 the same months. The character of the voltages varied widely from sharp 

 transients of a few millisecond duration to slowly changing voltages lasting 

 .2 seconds from one zero value to the next, voltages due to multiple dis- 

 charges being quite common, the interval between voltage peaks in such 

 cases being in the order of .1 second. Of the disturbances, 90% lasted for 

 more than .1 second, 50% for more than .4 and 10% for more than 1.25 

 second, the maximum duration being 2.3 seconds. By way of comparison, 

 the observed duration of discharges to a tall structure (3) were, in respec- 

 tively 90, 50 and 10% of all cases, in excess of .08, .3 and .6 second, the maxi- 

 mum duration being 1.5 second. The maximum voltage recorded was 940 

 volts and was probably due to a stroke to ground near the cable. About 2% 

 of the voltages were in excess of 500 volts, most of these and the lower volt- 

 ages being due to discharges between clouds, as indicated by the opposite 

 polarity of the voltages at the two ends of the test section. The wave shape 

 of the voltages at the ends of the section were much the same as at inter- 

 mediate points, even for the sharpest surges recorded, the attenuation along 

 the core-sheath circuit being quite small. It is possible that substantially 

 higher voltages than observed may obtain in the case of severe discharges 

 along a path parallel to and directly above the cable, and that such voltages 

 may produce cable failure if the core insulation is below normal. 



While the oscillographs were arranged to trip on 100 volts, a smaller 

 voltage was recorded in 40% of all cases, as the peaks were too fast to be 

 recorded by the type of oscillograph used. It is also possible that for this 

 reason fast voltage peaks in excess of the maximum given above may have 

 escaped measurement. 



II. Lightning Trouble Expectancy 



2.1 General 



In estimating the liability of a cable to lightning damage, it is assumed 

 below that once the core insulation is punctured, as it is likely to be at several 

 points, at least one permanent failure will occur. The lightning trouble 

 expectancy curves presented here thus give the number of times lightning 

 damage is likely to occur, without consideration of the extent of the damage 

 on each occasion. Each case of lightning damage usually involves several 

 pairs and, based on experience, repair of each such case would require about 

 four sheath openings. Damage due both to direct strokes and strokes to 

 ground is included. Discharges between clouds have been neglected as a 



