Our ride-on railroads are outdoor railroads. In addition to being able to enjoy the outdoors, we have to contend with nature’s wrath. One of the things we need to deal with is lightning.

Damage caused by direct lightning strikes to any signal system is obvious. I don’t know of any ride-on railroads that can afford the kind of protection it takes to survive this type of event – even relays and toggle-switches will be destroyed.

The question is how do we protect ourselves as much as possible, do it at a reasonable cost, and do it without sacrificing functionality?

Signal systems based on relays and incandescent lamps do fairly well at surviving nearby lightning strikes. Most people attribute this to the fact that relays and bulbs are “more robust”. I attribute it to the fact that these type of systems present a higher load to the rails, wiring, etc. More on this below.

There is also a common misconception that solid state systems, while being able to provide more functionality (remote control of signal indications by a dispatcher, ability for dispatcher to see all occupancy without extra wire infrastructure, etc.), are inherently less robust. I would disagree with this, but to justify this we need to look into lightning a bit more.  (See also: The Viability of Solid-State Components for Signal Systems)

Lightning is a very complex, and seemingly random phenomenon.  There is a great deal that science does not yet understand about lightning.

The Major Causes Of Lightning Damage To Our Signal Systems

1. Direct strike – Fairly obvious.
2. Induced Voltage – The current of a strike (on the order of 100,000 amps) creates a large magnetic field that rapidly expands and collapses.  This field can “induce” (like in a transformer) large voltage spikes in any conductors (wires, rail, fences, etc.) within the magnetic field created by the strike.
3. Ground Voltage Gradient – A lightning strike creates a localized voltage on the order of 90,000 volts at the point of the strike.  As you get further away this voltage drops off at around 100 volts per meter (depending on soil conditions) out to about a kilometer or so (this equates to about 30 volts per foot).
4. Commercial Power Feed – This is a less obvious source of lightning surges, but, in my experience a significant one (see below).

The Major Effects Of Lightning Damage To Our Signal Systems

Direct Strikes

We’ll mostly ignore this one, all bets are off, we don’t have the necessary budgets to deal with these.

One thing that can be done within limited areas (main yard, etc.) to minimize this source of damage is to install lightning rods.

In general, a lightning rod will protect a cone-shaped area extending from the top of the rod down to the ground at a 45 degree angle.

A lightning rod serves two purposes:

• Some lightning rods have a sharp point on it which prevents the buildup of a localized charge (the sharp point dissipates the charge into the atmosphere).
• The rod is grounded with heavy braided cables to a heavy ground rod, or better, a “grid” of wires and rods buried in the ground.

To be fully protected, the installation needs to be done properly and you should consult a qualified lightning rod installer.

The downside of lightning rods is that they do not “get rid of the lightning”, they merely direct it into the local area around the ground rod(s) (see: Ground Voltage Gradient below).

Induced Voltages

As mentioned above, the very high current in a lightning strike causes large voltage spikes (multiple for each strike – at least one on initiation and one on completion of each strike) to be induced into commercial power lines, rails, cables in the right-of-way, etc.  The level of this induced voltage depends on many factors including:

1. The distance from the strike
2. Length of the conductor
3. Relative location and orientation of the conductor to the strike path
4. ‘Load’ on the conductor

Induced voltages cause voltage spikes on the rails, wires going to signal heads, and the power/data cables, all of which are directly connected to block detectors, controllers, signal heads, etc.

Ground Voltage Gradients

The effects of this are a little harder to grasp but I believe that they account for a large percentage of the problem as it effects ride-on railroad signal systems.

Assume lightning strikes ¼ mile away and in-line with two track-side devices (rail detectors, block controllers, signal heads, etc.) located 300 feet apart.  This means that the voltage in the ground at the two devices might have a difference of 9,000 volts (300 feet * 30 volts per foot).  If the two devices are connected in any way (power cable, data cable, rail, etc.), there will be a voltage difference of 9,000 volts across that connection – not good.

Each device might “see” 9,000 volts at it’s terminal (complex formula depending on the load at each end, etc.) and 9,000 volts is not friendly to solid state devices – or even relay contacts (burned or fused).

The signal device can protect itself if it shunts this voltage difference around itself to ground (provide a low resistance path) to prevent the surge from passing through the device itself.

The resulting surge current is limited by the resistance of the ground and other factors (inductance, capacitance, etc. which I won’t go into).  If the track-side device can handle the resulting fault current, it can protect itself.

I attribute the “robustness” of relay and lamp signal systems to the fact that they generally provide a larger “load” (lower resistance path) to the conductors and to the fact that they can handle short over-current and over-voltage events (obviously it depends on the magnitude and duration of the event).  Note: this assumes “incandescent” lamp systems, those using LEDs are more susceptible because LEDs are easily damaged by short, high voltage spikes.

I believe that a solid state signal system can be made as robust as a traditional relay based system by incorporating inexpensive, fast, over-voltage shunting devices on all connections to track-side devices.

This allows us to implement a signal system with more functionality than is possible with a relay system while retaining a high level of reliability.

Commercial Power

In my testing, I have noticed that the devices located closest to the signal system power supply have sometimes suffered the most damage.

I believe that signal systems suffer more from surges coming in on the commercial power feed because signal systems are usually in intimate contact with the ground.  Normal household appliances are usually “floating” so a “common-mode” surge (same surge on both power and neutral) are tolerated better.

In my testing I have noticed that surges can even come into the signal system on the “neutral” commercial power conductor.

My Approach To Lightning Protection

Commercial Power Connection

I recommend that signal system power supply NOT be connected to the “protective ground” provided by a wall outlet because this is not a low-inductance path to ground.  I recommend that a MiniRail Solutions Cable Interface Card be used at the power supply location and be connected to a nearby ground rod as described in the documentation.

I also recommend placing a two-pole relay or switch in the AC power connection to the signal system power supply such that it interrupts BOTH the power and neutral lines to the power supply.  This will provide an “air gap” to protect the power supply and signal system when the system is turned off (which is usually 90% of the time).

MiniRail Solutions Device Protection

My block controllers use surge suppression devices. These devices shunt any surges (power line, track circuit, etc.) to the local ground to protect the components within the controller. This protection may not be needed if wire runs are short (such as to a nearby signal head, etc.) or if the facility is located in an area where lightning strikes are uncommon.

I also employ fuses between each shunt device and it’s field connection. These fuses serve to protect the shunt device from a surge strong enough to destroy the device itself.

Unfortunately, the shunt devices have both a maximum current rating and a maximum power rating (current times ‘time’). What we want to do is use a fuse size that prevents them from blowing unnecessarily but still limits the surge to a level that does not destroy the surge protection devices. Since lightning is unpredictable, it will take some time for field testing to determine the best value for these fuses. Until then, I use the more conservative value which means that fuses will probably have to be replaced more frequently but the surge protection devices will be protected.

It may turn out that, for 99% of the cases, the fuses can be eliminated resulting in a cost savings as well as smaller track-side devices.

More field testing will tell …

In the lab I have drawn an arc to one of the controller Track Input terminals to test the surge protection.  The arc was approximately 3/16″ of an inch long.  I used an AC power supply that generated about 6,000 volts.  Since it was an AC supply, I was testing both positive and negative surges in relation to ground.  The controller suffered no ill effects and the fuse did not blow.