GG1s and Transformer Oil

2508 views
16 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    September 2020
  • 11 posts
GG1s and Transformer Oil
Posted by TrainsButSmall on Wednesday, October 28, 2020 3:20 PM

The GG1 might never run again due to the transformers being banned due to the oil containing PCBs. However, I have been wondering if you could substitute the PCBs with a different form of transformer oil (Such as silicone oil) and run a GG1 with the original transformer design save for the different oil. This would satisfy requirements for safety and preservation alike!

Technically, this would be the equivalent of converting a steam locomotive to oil burning for fire safety.

 

I get that there are almost no surviving transformers, but you could build a new one to the original design (excluding the oil).

 

Would this work?

  • Member since
    January 2019
  • From: Henrico, VA
  • 5,899 posts
Posted by Flintlock76 on Wednesday, October 28, 2020 4:19 PM

I wouldn't know bout how well your idea would work, but the last I heard of the "PCB's in the transformer cooling oil" controversy was whoever started it had an Emily Litella "Never mind!" moment, that is, the PCB content wasn't anywhere near as bad as initially reported.

What IS true, is that the GG1's still in service at the time were pretty ragged out after nearly fifty years of service, most developing frame cracks by that time.  

If you don't know or don't remember who Emily Litella was, Google "YouTube Gilda Radner as Emily Litella."  Good for some laughs anyway, she was one funny lady!

  • Member since
    September 2003
  • 14,761 posts
Posted by Overmod on Wednesday, October 28, 2020 5:00 PM

TrainsButSmall
I get that there are almost no surviving transformers, but you could build a new one to the original design (excluding the oil).   Would this work?

The answer is a bit complicated.

The original transformers were designed, and their very complicated  frame, winding/tapping, and insulating structure designed around, the use of Pyranol as the transformer 'oil'.  This had high-temperature, high-current advantages.  

https://steampoweredradio.com/pdf/general%20electric/components/general%20electric/General%20Electric%20Pyranol%20Transformers%20Including%20Pyranol%20Immersed%20Current%20Limiting%20Reactors%20Instructions.pdf

The stated reason for not replacing the 'fluid' was that it was technically impossible to remove much of the Pyranol, more precisely its degradation products (dioxin and furan), from the transformer internal structure without compromising it electrically.  This was not a decision reached lightly or 'solely by nanny-state bureaucrats'.

Theoretically yes, you could build a replacement transformer that would run a GG1 in excursion service.  It would still need to be oil-cooled.  I have not looked to see if modern replacements that perform as well as Pyranol have been developed in recent years, but designing to lower amperage requirements should allow equivalent tap structure, and I think there are clearly better internal insulation materials for actual fabrication.

https://www.usbr.gov/research/publications/download_product.cfm?id=2822

The principal issue is cost.  Building one of these is a complex, safety-critical operation ... and when you're done it would be restricted to 25Hz and no more than 12.5kV, ruling out both most mainline operation and cost-effective museum operation at the same time.

Redesign of the main transformer to 25kV is an obvious improvement; detail design could probably be made through nothing more complicated than a good college engineering contest backed up by expert design analysis.  But here again, construction of the new machine followed by careful rebuilding of the rest of the running gear would likely involve multiple millions of dollars, followed by expensive regular maintenance.

As some have pointed out, it would be cheaper to adapt other components, one example being Mr. Klepper's adaptation of AEM-7 electrics to supply filtered DC to the GG1's universal twin motors.  (The problem there being that Amtrak, for somewhat misguided liability reasons, does not allow sale of any of its locomotives for subsequent operation; they went so far in a couple of the early scrappings to cut the wheelset axles and side frames to rule out their being adapted to any operable purpose...)

Note also that the New Haven prototype from which the GG1 was taken had arrangements to use NYC 750VDC third-rail power from New Rochelle via Woodlawn, far more than just 'accommodation' to negotiate the Park Avenue viaduct at its restricted speed.  So logically any museum with 600V trolley power and overhead suitable for pans could arrange for a G to be, comparatively easily, 'reworked' to motor a couple of axles without grossly overloading things...

  • Member since
    January 2001
  • From: Atlanta
  • 11,751 posts
Posted by oltmannd on Thursday, October 29, 2020 9:01 AM

When Pyranol was banned, Conrail, NJT, MNRR, LIRR et.al. replaced it everywhere they had it with mineral oil.  Not just in the transformers in locomotives and MU cars, but in wayside transformers as well.  

There were two problems with this.  One is that mineral oil is flamable.  The other is that the Pyranol had leached into the insulation material and slowly leached back out into the mineral oil.  So, everytime you had a spill of any kind, it was still treated as hazmat.   Folks had to keep testing the mineral oil to find out what the concentration of PCBs remained.

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    September 2003
  • 14,761 posts
Posted by Overmod on Thursday, October 29, 2020 9:53 AM

oltmannd
There were two problems with this.

There was a third problem, which I think wisely went somewhat unreported: While PCBs are 'flame retardant' they will certainly support combustion at not too high a temperature (in the 300s C) and sustain it not too far above that temperature.  The products of combustion are NOT pleasant, some at concentrations of a fractional ppb (yes, billion).  The substitution of mineral oil immediately raised the probability that transformer fires would produce precisely the kind of contamination that (rightly, in my opinion!) gave dioxins such a bad reputation.

If this was not directly related to the sudden demise of the E40s I'll be surprised.  To my knowledge those transformers were still paper-insulated and oh, brother! were they effectively undercooled for what was expected of them in PC service!

  • Member since
    September 2020
  • 11 posts
Posted by TrainsButSmall on Tuesday, November 17, 2020 3:55 PM

 

Some railway museums do own trolleys. Perhaps those could run the 600v GG1s

  • Member since
    September 2003
  • 14,761 posts
Posted by Overmod on Tuesday, November 17, 2020 6:50 PM

TrainsButSmall
Some railway museums do own trolleys. Perhaps those could run the 600v GG1s

A GG1 weighs over 220 tons all by itself.  So for most "trolley" museums, their 600V substation or power source would probably require significant upgrading even if only minimal consists were handled at comparatively low speed.

  • Member since
    January 2019
  • 811 posts
Posted by Erik_Mag on Wednesday, November 18, 2020 4:17 PM

IIRC, the rectifier substation at OERM is good for maybe 600kW continuous and perhaps 1200 to 1500 kW for short term. This would likely support series parallel operation with two sets of six motors in series.

  • Member since
    December 2007
  • From: Georgia USA SW of Atlanta
  • 10,057 posts
Posted by blue streak 1 on Friday, November 20, 2020 11:28 PM

PCBs in transformers are a gift that just keeps giving.    Our local town has approximately 25 - 35 transformers still with PCBs,    The problem is mostly ignored but when hurricane Opal crashed the transformer outside my home PCB was spilled.  The clean up?  Believe it or not city cut up trees and used the saw dust to asorb the oil.  Where it went have no idea. 

  • Member since
    September 2020
  • 11 posts
Posted by TrainsButSmall on Sunday, November 29, 2020 1:30 AM

I agree, I have volunteered there, but the problem is, we're in California. This is a Penn Central / PRR Locomotive.

 

Are there any other such museums with substations capable of this or are OERM's Del Paso substation and the newer substation at OERM the only possible GG1 candidates?

  • Member since
    January 2019
  • 811 posts
Posted by Erik_Mag on Sunday, November 29, 2020 11:59 AM

Keep in mind that the GG1 would have to go some major electrical work to make it operate on 600V. FWIW, one reason that OERM was not interested in acquiring a GG1 was the problem with dealing with the PCB's.

The museum at Rio Vista Junction has a solid state substation that was originally used by BART for testing along with a portable substation.

  • Member since
    September 2003
  • 14,761 posts
Posted by Overmod on Sunday, November 29, 2020 2:05 PM

Erik_Mag
Keep in mind that the GG1 would have to go some major electrical work to make it operate on 600V.

Not sure it's all that much -- assuming it's just for museum operation at comparatively low speed.

His first step should be to find wiring diagrams and specs for the New Haven passenger motors that used this chassis arrangement (and operated on 750VDC, a higher tension, into Grand Central with heavy trains).  While I think these were out of service when Joe started training, he may have some knowledge of how the DC and AC controller functions were implemented.  I do think it is safe to ASSume that we don't need to care about either continued or future "AC capability" if a museum were to do such a conversion, so much of the existing motor and internal wiring might be 'adaptively reused'.

If only one of the twin-motors were used, I suspect a simple trolley contactor could be used (with comparable electrical components); I'm too lazy to do actual calculations as to what would safely run this type of load, or how to wire ground-fault detection or if the existing overload breakers etc. would serve on DC.

  • Member since
    January 2019
  • 811 posts
Posted by Erik_Mag on Sunday, November 29, 2020 6:50 PM

I would go the solid state route. Cree/Wolfspeed has a half-bridge module capable of handling 700+ Amps with a 95C baseplate temperature using 1200V FET's. A minor challenge is to brew up the software for controlling this and the whole package would be much smaller than the switchgear used on the GG1. IIRC, the rated terminal voltage for the motors is ~225V.

  • Member since
    September 2003
  • 14,761 posts
Posted by Overmod on Sunday, November 29, 2020 8:03 PM

Erik_Mag
IIRC, the rated terminal voltage for the motors is ~225V.

Now that I did not know.  I thought they were insulated far higher than that ... well, then again, they'd have to be!  (Bet there are still people who remember the fix for the '58 'diamond snow' and what electrical rating the epoxy was good for.

220V is an interesting reference point for potential mobile power sources, as well as control drives.  Bet it wouldn't be hard to find up an API for touchscreen function control... and integrate it with 'other information'... Wink

For some reason I remember ASEA did an exhaustive review on how to remotor a GG1 -- but it was all high-voltage AC based.  All the stuff I saw regarding rebuilt GG1s for high speed in the Corridor involved AC motors as well, and it would have been fun to see at least one test conversion done... although it well might have been a late-70s precursor of the Republic Starships and there really was no suitable way to implement sensible brakes for the intended service...

  • Member since
    January 2019
  • 811 posts
Posted by Erik_Mag on Sunday, November 29, 2020 10:01 PM

AC series motors are a compromise solution, they;re larger, less efficient and a bit more cantankerous than a DC series motor. OTOH, up to ca 1950, they were the only way to get variable speed without an onboard M-G set from a single-phase electrification.

One of the compromises is using single turn windings on the armature and that is what limits the terminal voltage. It is quite likely that the two motors per axle are wired in series to reduce the amount of copper needed for the motor leads.

As for the Cree/Wolfspeed module, I was proposing to use PWM to provide a variable DC voltage (AKA chopper). With a 20kHz PWM frequency, it would be relatively simple to provide pure DC to the traction motors. It also would not have the duty cycle limiations of Jones choppers implemented with SCR's.

  • Member since
    September 2003
  • 14,761 posts
Posted by Overmod on Sunday, November 29, 2020 10:31 PM

CAB760M12HM3 shows a recovery time of under 50ns, which ought to allow high-frequency PWM drive at under the nominal <20V.  Might want to liquid-cool it, just to be safe...

  • Member since
    January 2019
  • 811 posts
Posted by Erik_Mag on Monday, November 30, 2020 10:12 AM

That's the module I was referring to and liquid cooling would be mandatory to get anywhere close to rated current. IIRC, there should be COTS cold plates that fit this module. Limiting factor on switching frequency is turn-on and turn-off energy.  The rule of thumb is that switching losses should be roughly equal to conduction losses.

This module is intended for use in traction inverters running on a 800VDC bus, where three would be needed. OTOH, it has about the right ratings for doing a solid state replacement for a PCC motor controller.

 

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy