Electrical Grounding in Real Railroading

Prr7688 posted:

Hi everyone, I'm an electrical engineering student getting ready to graduate (sadly my grad ceremony has been postponed, thanks corona) and have a job lined up in the power engineering industry. I was wondering about safety requirements surrounding electricity in the railroad industry, mostly about grounding of equipment, and came up with a few questions. Any input would be great: I'm curious to learn more on the topic! 

For steam locomotives: is the negative side of the dynamo connected to the metal parts of the locomotive thus connecting it to ground through the rail?

No. The electrical system in steam locomotives was DC, with a positive and negative wire. Absolutely NOTHING was grounded.

In short lines running mostly steam: how often is the track grounded? (i.e. how many feet between grounds rods in the earth?)

Never saw ANY track grounded!

How does that distance change for Class I railroads?

Except for electrified railroads, i.e. operating with electrified overhead or 3rd rail, there is no grounding.

How does that change for areas running electric traction? (I do know that the tracks are grounded at the substations along the line but there could still potentially a high fault current through the rail until the breaker trips and figure there are earth grounds)

Does jointed rail vs. continuous welded rail make a difference in any of these numbers? 

The signal systems use a potential between the two rails to detect the presence of the train. It is important that the rails are not be grounded. 

Electrified railroads use the the rails as the return conductor. The rails are not grounded. This is important to avoid electrolysis damage on parallel underground utilities.  The potential between the rail and ground is monitored at substations, but is only connected if potential difference gets to high. I recall an IEEE paper where voltages were measured between the catenary towers and the running rail on the north east corridor.  Voltages in excess of 1000 volts were measured when heavy trains were accelerating in the area. I did clearance tests using aluminum fingers on the outside of a subway car. At the ends of the platforms these is a short metal fence which is in side the clearance envelope. When the metal fingers touched these fences while the train was accelerating out of the station there was considerable arcing. I was told the voltage potential was only about 40 volts. 

On jointed rail the joints are bonded for both signals and electric traction. The size of the bond depends on the potential current in the rail.  On DC electric railroads with ac signal systems, impedance bonds are used to isolate the signal blocks from the traction power.   This is a complex issue with a lot of people studying it.

Wow! I never imagined that grounding wasn't that much of a concern. The reason I figured it was is the grounding regulations surrounding utility lines. For utility lines, poles have to be grounded at regular periods so that a fault current doesn't travel through the surface of the earth that far before being effectively grounded. I figured that there would be a similar grounding system with rails so that there isn't a fault current through the rail.

It makes sense that the signaling systems prevents the ability for rail to be grounded, in my mind there was a way for there to be both grounding and signaling. 

David, thanks for mentioning the IEEE paper! It sounds really interesting, I'm going to try to find it online. 

Prr7688 posted:

Wow! I never imagined that grounding wasn't that much of a concern. The reason I figured it was is the grounding regulations surrounding utility lines. For utility lines, poles have to be grounded at regular periods so that a fault current doesn't travel through the surface of the earth that far before being effectively grounded. I figured that there would be a similar grounding system with rails so that there isn't a fault current through the rail.

It makes sense that the signaling systems prevents the ability for rail to be grounded, in my mind there was a way for there to be both grounding and signaling. 

David, thanks for mentioning the IEEE paper! It sounds really interesting, I'm going to try to find it online. 

For what it's worth, "grounding" in the railroad industry is NOTHING like "grounding" in the power generation industry. For example, even diesel electric locomotives, from 1934 through today, do NOT have a grounding system, as the low voltage control circuits are ALL DC, i.e positive and negative wiring, and the high voltage power distribution systems (both DC traction and AC traction) are all isolated systems. Each and every diesel electric locomotive is equipped with a ground fault detection system, which removes all high voltage electrical power very, VERY quickly, by unloading the prime mover.

PRR 7688 

As Hot water notes above AC systems are most definitely isolated as well as the DC traction motor Power Circuits. I would think the AC traction systems would be a closed Delta system for the traction motors as well as the train lines (power distribution). The gensets, I believe would be grounded to the frames. The Acela's shore power/ train line is 480V, 400A, 3 phase delta. Train power and commercial power are two different animals. I have been watching AC's migration into the railroad world.

On our 70 year old 44 and 45 ton locomotives we have both positive and negative grounded locomotives. The GKK lead off the main generators are grounded to the frames. the 44 ton is negative ground and the 45 ton is positive ground. The batteries also match the genset. The 45 ton's positive lead is grounded to the frame at the main disconnect. the 44 ton negative is grounded to the frame at the main disconnect. So you can have both.

There are about 8 sub ground points through out the locomotive to ground the control circuits. we feed out normal lighting and control circuits and then ground them to the car body frame, similar to an automobile.   

Now tell me why when we double head the two locomotives sparks aren't flying??? open one of the main ground jumpers and I'm sure they will.  

Joint rail has issues and the RR's add bonding jumpers at every joint in signalized territory via Chicken necks bonded to the rail. The rail tampers come along and cut them right off if there not tucked in real tight to the joint bar. ( I personally know!)The rails need to be isolated for signals to work, and need to reduce the track resistance as much as possible. The rail wheels short the two rails together sending the signals (shunts) back to the control cabinets. due to there light weight Speeders and Hi Rail's have there axles isolated so not to shunt the rails. Hi rails have built in shunt switches to activate signals at crossing when needed, otherwise as these machines bounce up and down the rails, they would be driving the signal equipment crazy.  

I'm pretty sure all metal framing and structures are well grounded near a catenary system. Both for safety and cathodic protection. How the grounds are set up for the ground return on catenary systems I have never looked into. Single point I would imagine, with all segments bonded together. The left and right rail need to be insulated from each other though. so do they only ground one rail, i would imagine so.

I saw a paper explaining this and don't remember where it is.  It explains how the PRR could run the NE corridor with the HV train sets via the catenary systems, train power,  Run cab controls, signals, crossing gates, radio phone systems   All using the same piece of track.  

Do signals have some sort of grounding to protect from lighting strikes?

When I was Assistant Superintendent at Sweetwater, TX, I wore a lot of hats, when needed, as we were not a major terminal.

Our Communication Technicians were located at San Angelo (75 miles south) and Brownwood (110 miles southeast).  They shared the territory between Sweetwater and Temple, and, when we had big thunderstorms, were spread thin.  The San Angelo fellow asked if I would be willing to be his emergency backup, for restoring power to an important 2-way radio antenna which sat on a bluff near Buffalo Gap.  Of course, I agreed, and so we set up a day to have him take me up and show me what to do.  We had to cross two ranches, using gates that had 4 padlocks each, drive on a turkey trail up onto the bluff, and then enter the bungalow.  The antenna had a very large braided lightning arrestor wire with an in-line fuse about the size of a fusee, that protected the radio electronic equipment.  He showed me where the supply of fuses was kept, and how to re-start the radio equipment.  We went to lunch and he gave me a set of keys.  

I had to go up there twice when we had big storms, and then, of course, the turkey trail to the bluff was all mud.  But that is the normal life of a Communications Technician on the railroad.  They don't get much thanks for doing a rough job that almost always has them out when the weather "gets western."  I made a point of taking him to lunch once every month or 6 weeks after that, and did the same for our two Signal Maintainers.  They were always there in bad weather, and did not whine about it.  

I'll post a story on another thread about our Brownwood Communications Technician.

David Johnston posted:

Electrified railroads use the the rails as the return conductor. The rails are not grounded. This is important to avoid electrolysis damage on parallel underground utilities.  

As a side note, this is one reason streetcar systems removing some, but not all, of their rail lines often paved over rail rather than removing it, as they found out that removing the ground that the rails provided would increase electrolysis problems with underground utilities... 

Mitch 

Seems like in this thread thete are two types of grounds:

1.  True earth grounds.

2.  Common grounds, like the use of a carbody on a auto or truck, or the outer two rails on 3R track.

Are there more?  And are there correct terms for them?

There are term definitions of both ground and bond.  A large part of the NEC (National Electrical Code) is about grounding/bonding.   Seems like, over 40 year this subject was constantly changing, being upgraded.    I had done ground grids for several cell tower sites, years ago.  There was a potential test, done on the grids, once completed.  Lower the number the better.   The common, one, or two 8ft. rods, (residential service) doesn't do a lot, though with plastic water lines both inside and outside homes, a once reliable metallic water system is no longer available, may be creating additional problems. 

Our house in the South Hills of Pittsburgh was beside the T.:

Our water line developed a leak and was flooding our neighbor's driveway so I called the Plumber. He dug up the [copper] line and replaced a section. I remarked on the deep pitting of the removed section and he pointed to the T tracks, saying their lack of grounding caused electrolysis and eventual failure of our copper water line. Quite common according to him.

When the Houston METRO LRT Red Line went in UH Downtown to Fannin South, there were problems with keeping current in the ROW.  It was fixed.  But a big draw of the leaky current was the Texas Medical Center's huge steam generating plant about 3 blocks 3 of the TMC TC....

Did the 4 rail system of the London Underground stop leaks?  2 running rails, one Pos and one Neg rail?

M. Mitchell Marmel posted:

David Johnston posted:

Electrified railroads use the the rails as the return conductor. The rails are not grounded. This is important to avoid electrolysis damage on parallel underground utilities.  

As a side note, this is one reason streetcar systems removing some, but not all, of their rail lines often paved over rail rather than removing it, as they found out that removing the ground that the rails provided would increase electrolysis problems with underground utilities... 

Mitch 

Some commercial ovens use ground as part of a safety to keep the gas valve powered for the oven to heat. In Chicago some of the street car rails still buried are now DEEP under the current street level. Over the years I’ve been involved a few times where the city has dug up a street for utility work, and removed some old street car line rails in the process. It’s insane trying to explain this whole story about ground issues to pizza place owners, service companies, and then an electrician or power company.