Fun with trains.

[itsmoked]

We have two rail cars that run happily on their own generators. This costs money for fuel that is otherwise provided by HEP as part of the hauling deal so it is VERY desirable to use HEP when it's available. If either of these cars are put up against the engines the entire hookup described above works fine but the main HEP contactor trips on "fault". If the car is placed at the back of the train there is usually NO problem.

These cars consist of HEP delivered thru a 480V, 50A breaker to a bus-bar system. A 480V - 240V transformer bank,(3 dry transformers), then feeds all the 3 phase loads and single phase loads like ovens. The 240V is transformed down to 120V to feed single phase loads.

I'm looking for suggestions on why these cars are tripping out the HEP.


################### Ancillary Info ################

H.E.P. = Head End Power

HEP is power provided by a locomotive to Amtrak passenger cars. It is passed down four huge cable sets. The plugs are about 60 pounds apiece.

Each of the cable sets has 6 conductors. One for each phase (Delta) and then three safety wires. To prevent arc-flash casualties the pins are staged. The three safety pins engage last and disengage first.

The power is 480V @ 1,400A split across these four cable sets. Once all the cables are correctly engaged a contact closes in the engine that turns on a light signifying that the main contactor can close. The engineer can then press a button that closes the contactor that energizes all four HEP cables. Any pin disengagement automatically drops the HEP.

The cables run in pairs down both sides of all cars. 72VDC is sent down the three safety wires of one cable and at the end of the train are "looped back" with a full cable jumper returning down the adjacent three safety wires of the neighboring cable. These three signals are ANDed together at the engine to confirm two completely connected HEP circuits on each side of the train.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[davidbeach]

Keith, nice description of the HEP system, but what is the "fault"? Is it loss of continuity in the safety conductors or is it some type of electrical fault on the 480V system. More information needed.

 

[jghrist]

If the 480 volt system is delta, is it ungrounded? Is there ground fault protection/detection? Any indication of the current during the "fault"?

 

[waross]

My first reaction is that the starting current of an A/C unit is drawing enough current to trip the HEP breaker. Possibly on instantaneous, possibly on time/over-current. The added impedance of a train length of conductors may limit the peak current enough to prevent instantaneous trips.
Does the hep trip on energization? Possibly more than one A/Cs are trying to start simultaneously.
Can you design some hand wound series inductors to limit reduce the peak inrush. A reactor that gives a 1 volt drop at FLA will possibly give dramatic results.
A simple calculation would indicate that under starting current the voltage drop may be 6 to 8 times the normal running voltage drop. However the reactive nature of the starting current may result in a much greater voltage drop under starting conditions than expected.
It is probably that troublesome 'first quarter cycle' energization surge that is the culprit. It is possible that even much less than a 1 volt drop at FLA will be enough to avoid breaker trips.
What is the impedance of a train length of cables? Probably less than an ohm. That value of reactance in series with your equipment may avoid your trips, and due to its quadrature nature be almost un-noticable at FLA.
I don't have time to crunch the numbers, if you do, let us know how they look.
Respectfully
Bill

 

[itsmoked]

Thanks guys.

davidbeach; No it is NOT a loss of continuity problem but some sort of fault.

jghrist;(See below). (See below). No current level indication is available.

Clarification: It is the 1400A breaker opening after the contactor closes.




OK, after more grilling:

The system shows "[green]ready[/green]". The "close main contactor" button is pushed => bang-BANG the main breaker trips instantly.

There is a ground fault safety system. The rails are used for the ground with large conductors run to the trucks/wheels -> rails. There is no [red]red[/red] light Ground Fault indication ever.

All the Amtrak cars have single phase 480V heating systems with various floors heated on various phases and that conceivably all of them could have only their first floor heaters ON and this doesn't cause any problems.

Lastly these two cars have been checked for load symmetry by Amtrak. It is less than 5% or the cars are not allowed to be hooked into the grid in the first place.

If these two cars are hooked up at the back of the train together the problem persists. If they are hooked up at the back of the train with a few,(more than one), Amtrak cars between them then the problem will likely not manifest itself.


I am thinking Bill is on to something. I am still scratching my head because our cars have nowhere near the loads the Amtrak cars have. Single deck Amtrak cars are 50A, 480V. Double deckers are 100A.

Our cars run everything on about 50kW. Yet 480x50x3=72kW.

Other differences... We have NO 480V loads. All refrigeration is 240V. All heating, all cooking is 240V. Amtrak runs all refrigeration directly on 480V. This means we have bigger transformers.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[davidbeach]

Keith, same cars making up the rest of the train? If so, I don't think that Bill has it as the order would not make a difference unless you get upward of 20 cars or more (length of circuit).

Check your cars end to end with a milli-ohm or micro-ohm meter - one of those used for measuring breaker contact resistance. If you have a high resistance connection somewhere in your car and the rest of the train is essentially a constant kW load, the voltage drop across your car will increase the current in the train circuit as the other cars try to compensate for the decreased voltage. When you are at the end of the train it would only be your cars affected by the high resistance connection. Perhaps not even than depending on where the bad connection is relative to your loads.

 

[JensenDrive]

Maybe a contactor is not closing reliably, and you intermittently are single phasing some three phase loads, causing high currents on two legs.

Are you saying you have 3phase 240VAC loads/refrigeration? Or is this a set of single phase 240 loads that appear as a balanced three phase load only when all thermostats are on? Can unbalance current flow cause a trip?

Sound like you need to devise a way to measure voltages and currents during the incident.

The equation for 3ph using balanced line to line voltages and balanced three phase current is
VA = Sqrt(3) * Vll * I
VA = Sqrt(3) * 480Vll * 50A = 41.6kVA

 

[itsmoked]

davidbeach; The Amtrak cars are NOT the same, as they are newer and 'engineered'. These cars were Amtrak decades ago and are now, pretty much, entirely custom, in some cases having come over from DC service.

A train is often 20 or more cars.

Yikes.! I'll have to think about making a milliohm reading over a 90ft distance. Hmmm maybe I can loopback and keep the measurement at one end. How sensitive are we talking on any resistance? These plugs are huge mal-treated devices. I have seen them heated with torches so they could be plugged together after being fished out of the snow.

Our HEP systems are, in these two cases, rather newer than the Amtrak ones generally. Also why would both our cars have a high resistance connector problem? We use the same Amtrak custom hardware.

JensenDrive; DOH! Thanks for the reminder on the Equation.
I'm having problems with a faulty contactor theory too. This is because both cars behave exactly the same. Two faulty contactors?

Our 240V loads are either heater elements that are generally hooked up in delta,(unlike Amtrak's 480V heaters), or are 3ph motors.

There is an exception to this! There is a transformer on one pair of the 240 legs to create the 120V that runs domestic stuff; lights, refrigerators, appliances.

These cars have three single phase 25KVA(?) dry transformers hooked up into a delta across the 480V and the same thing has been done to their secondaries to make 240V, 3phase (not 208V). I keep wondering about something possibly problematic in this arrangement. The Amtrak transformers are probably << 10kVA.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[davidbeach]

Keith, if both of your cars stay together, the high resistance connection only needs to exist once. The milliohm measurement is probably a four wire measurement to begin with and would absolutely have to be be four wire over the length of a passenger car. But you could plug in a jumper at the far end of the car and measure two of the four runs simultaneously.

If the contactor is in the locomotive, as you describe it to be, I'd tend to discount it as I don't see car order making a difference.

If it could be done, an interesting test would be to have the train assembled and a recording power meter (rpm or equivalent) in place on the locomotive. Then connect at one end of the train and energize the HEP, disconnect, move to the other end and reconnect and energize. Then compare the readings. Recording meters at each end of each of your cars would add useful information.

 

[waross]

Measuring milli-ohms, push 10 amps through the cable and go point to point with a volt meter. A 10 milli-volt drop equals 1 milli-ohm between the probes. Need more resolution, use more amps.
Relatively insensitive to lead length (within reason).
Try a few ohms in series with the transformer. Alternatively use fast acting fuses or a scope and energize from the gen-set to get an indication of the energization surge.
Can you switch in the transformer after the HEP breaker closes if only for a test? The load surge of a train load of cold heaters combined with the transformer inrush may be too much if the instantaneous trips are set at minimum on the HEP breaker.
I can't see a high resistance in a feeder causing an increase in the current drawn by resistance heaters.
Respectfully
Bill

 

[davidbeach]

Bill, you're absolutely correct that if the bulk of the load were resistance heat that a high resistance connection couldn't cause the problem. On the other hand, if the bulk of the load is motors and fluorescent ballasts then reduced voltage means more current. I'd still like to know what the trip condition is, setting vs actual current. Any solution has to account for the front of the train vs rear of train issue.

Keith, if you were to connect just your two cars directly to the locomotive would the HEP trip on energization?

 

[waross]

Three possibilities;
1 H.E.P breaker tripping on ground fault.
2 H.E.P. breaker tripping on instantaneous over current.
3. High resistance connection.

1 and 2, How much access are you allowed to the head end?
Are you able to determine if the problem is a ground trip or a current trip?

Re the 25 KVA transformers. I make the FLC = 90 amps. That times 2 may develop enough of an energization surge to trip the breaker. It may also saturate a CT and give a spurious ground trip, particularly if metering class CTs were used instead of protection class CTs. Scenario; the surge may be higher on one phase and that CT may saturate.

Suggestion; Are you able to disconnect one transformer (preferably both primary and secondary, one lead on each) and try an open delta connection? That should reduce the surge and may help point to the source of the problem.

3 I would use an electric kettle or hot plate to approximately 10 amps through the cable. Then check the voltage drop in the total cable. Calculate the resistance. If it is excessive, then do daisy chain voltage checks down the cable to see if the resistance is distributed or concentrated in a possible fault.
David, I accept your point. Sorry for not reading your post more closely.
Respectfully
Bill

 

[LionelHutz]

An immediate trip of the breaker tells me it's one of 2 problems.

1 - Instantaneous trip.
A breaker usually is thermal magnetic. An overload will trip after some time. A large overcurrent will trip instantly.


2 - Protection trip.
Some form of external protection is sending a trip signal to the breaker immediately. Does the loop detection with the safety wires trip the breaker or drop out the contactor? Maybe your wiring is different somehow compared to the other cars and the safety circuit is tripping it back off for some reason. Induced voltage in the safety loop of some sort??

 

[LionelHutz]

I should have expanded on #1 and said that the instantaneous overcurrent of most breakers is typically 4 to 16 times the rated current. So, your car would have to be drawing something like a 6kA surge to trip the breaker on instantaneous. You should try to find out what the breaker is and how it is set up.