BART Gremlins

[SD70M-2Dude]

Mysterious power surges have been disrupting service.  I linked to /. instead of the article it links to because their comments section is usually better than the actual article.

https://tech.slashdot.org/story/16/03/17/2310229/whats-frying-the-electrical-systems-on-bart-trains

The employees seem to have their own, logical theory on the root cause of the problem:

http://www.wired.com/2016/03/barts-righteous-tweetstorm-reminds-us-problems-fault/

https://twitter.com/SFBART

[Electroliner 1935]

If only more organizations had the will to try this. Voltage spikes on a DC system seem difficult to diagnose. It typically has a source from a utility line that could have a voltage issue on the AC side of the substation but unless there is a stuck tapchanger on a transformer or some capacitor banks switching on and off, these do not normally cause issues. On the DC side of the substation, which is typically just a unidrectional rectifier, the only thing I can see causing a voltage spike might be a train in dynamic braking trying to dump energy back into the DC system and not enough load to absorb it. Don't know how high the braking sytems voltage can go but I would expect that it is unlikely to be at a damaging level. As Yul Brynner said "Tis a puzzlement" Looks like they need some data recording voltage monitors. 

 

[erikem]

What struck me was that the cars being hurt are equpped with DC traction motors, presumably with some sort of chopper control and that the cars with induction motor drive weren't being affected. A couple of things come to mind - one is that the surge capability of the inveters is greater than for the choppers, the second is that the inverters need more local bypassing (i.e. capacitance) than the choppers and the local bypassing is taking care of the surges.

The "A" and "B" cars originally came with chopper controlled DC traction motors, with the motors connected as two sets of two motors connected in series (each motor rated at 550VDC). The chopper used ordinary inverter grade SCR's (no GTO) and provided for regenerative braking. The propulsion system also had a set of resistors that would be connected across the 1000VDC bus from the third rail when the bus voltage rose above 1050 to 1200V. (This is what I remember about a presentation given by a couple of BART engineers at UCB somewhere in the 1976 time frame.) Presumably, the "C" cars use a similar system with GTO thyristors.

I agree with Electroliner about putting on some data logging equipment, on the substations as well as some of the cars. It would be helpful to have some sort of accurate time information on the logs.

 - Erik

[blue streak 1]

BART temporary service but have found one section between stations that is having intermittent voltage spikes.  The use of just certain cars on the line may indicate a certain type car can handle the voltage spikes ?

http://www.bart.gov/news/articles/2016/news20160316-0

[MikeF90]

Just came across another interesting speculation involving substation transformers:

http://www.altamontpress.com/discussion/read.php?1,131142,131159

Still no 'new' news from BART, bus bridge is still in service.

[blue streak 1]

MikeF90

http://www.altamontpress.com/discussion/read.php?1,131142,131159

The linked article has a lot of merit.  Inadequate transformer iron cores for amount of current called for will cause many problems and will eventually fail.  This is somewhat related to the 25 - 60 Hz transformer problems and limiations on the NEC.

One more example of can happen with inadequate bid specifications and the demands to beat the other low bidders.

Interesting that this route is also having a much higher demand ( more cars & trains ) than forecast.

Also remember the WMATA transformer fire last year.  Sounds that all the traction power agencies need to look at their transformer capacities especially what happens if there is a bunching up of trains.  Also modern traction and electrical draw per car is higher now than in the past. 

[erikem]

I ran across an article stating that the spikes topped out at 2kV, twice the normal 1kV third rail voltage. The same article mentioned that the thyristors (SCR's) were about a grand apiece and a 20 week lead time from Powerex.

A bit of further research stated that the C car propulsion system is a copy of the original A & B car propulsion system, whereas the A & B cars have been updated with an IGBT based inverters supply induction motors. The inverter requires quite a bit of local decoupling/bypassing and that's probably why the A & B cars are not being hit by the problem.

A couple of questions do come to mind. #1 Are the C cars run in trains with A or B cars and if so are the cars that got damaged on a mixed train? #2 Were the new substations design to supply SCR choppers?