DC traction power feeder cable testing

Question

I work for a transportation company. And we deal with 650-800 V traction power. Mainly to third rail and overhead trolley wire.

We do a test on our underground cable. To basically see if it's becoming grounded, or is a ground. We have lots water problems in our manholes, and have been told that's why we have high results. The test is as follows:

Using an analog volt meter, red lead (+) on known hot 650 V source. Black lead (-) on out of service cable being tested. Usually we like to see 150 V or less. But we will most likely never see close to 0. I've been told that is water in the cable. Some areas we put it back in with 500-600 V results, and it doesn't trip the breaker. However it's not lighting a lamp bank, Which means it's not a good clean ground. (Lamp bank test is bulbs in series, connected same way from hot source to the cable being tested.

What is this test called? We call it a ground test, but I can't find that anywhere?

How come we test, and get 500-600 V to ground but it doesn't short and trip the breaker? Because it's not a clean ground correct? There would be resistance in the path to ground, correct?

How can we then get results of 50, 100, 200 volts to ground. If there's a ground at all, shouldn't it show up as 650 V?

Answer 1

Your analog meter will consume very little current (otherwise it'd get hot) so it only takes a small amount of leakage current to show up as a voltage difference. The lamp box requires a lot more current, so won't light unless the leakage resistance is low, and as for the breaker, it'd basically take a dead short to trip that.

You need to decide what leakage current or resistance is acceptable, then add a load across the meter terminals to put a sensible load on the leaky line; then if a significant voltage still shows up, the line is genuinely leaky.

Obviously with traction power you need to take suitable precautions in terms of insulating the connections to an appropriate standard, making sure the added resistor can handle the power dissipated if the load is connected across the full supply voltage, and protecting against serious injury from arcing if the supply is accidentally short-circuited.

I have worked with traction power on rapid-transit vehicles, and personally I'd strongly advocate using a 'Megger' insulation-tester when the power has been removed, rather than working on 'hot' cables, due to the serious risks involved.