Regenerative braking: basic-level questions

Thanks, Allen. It would be interesting to know one way or the other.

I am vaguely of the impression that the Y-1 was setup to automatically move between motoring and regeneration without switching or control intervention, the speed point at which the changeover occurred depending upon controller position. But I do not remember where I saw this – it was a long time ago – and I cannot trace it in any of the materials on-hand. So it may well be a distorted or inaccurate memory. But if so, then some alterations may have been needed to lockout regeneration capability.

Historical information in the compilation “Westinghouse Electric Railway Transportation” (which I did manage to find in the stack) indicates that the PRR power supply was a mix of dedicated single-phase turbo-alternators and motor-alternator frequency changers. In principle the latter would have been able to handle regeneration, although whether they were set up for “reverse flow” metering and whether the utility company involved would have accepted a reverse flow is another issue.

Cheers,

Pneudyne--
So far, no response on the PRR forum, alas. If, as think is pretty clearly the case, PRR hadn't used regenerative braking in the years prior to its acquisition of the Y-1 (FF-2), I doubt they would have seen the point to modifying their fixed plant to accommodate a small number o units with an anticipated short lifespan.
But maybe somebody who knows will eventually answer!

It’s a relatively obscure piece of information that we’re seeking, so I suspect that the answer would not be known to very many.

In search of possible pointers I did retrieve Brian Solomon’s “Electric Locomotives” book from the stackroom (aka garage). It has nothing on the Pennsy FF2. But it did mention that the earlier FF1 prototype was built in anticipation of electrification over the mountains to Pittsburgh, and that it was of the split-phase type with three-phase induction motors. Such are usually automatically regenerative, but Solomon said nothing in this direction.

On the basis of “you never know”, further searching was done to retrieve Staufer’s “Pennsy Power”, which turned out to be near the bottom of the last stack....

In respect of the FF2, Staufer noted that 48 separate structural, mechanical and electrical changes were made to adapt them to Pennsy practice, but no further details were provided. One of those 48 changes may well have been to isolate the regenerative braking facility. Staufer said that they were used mainly in pusher service, also in helper service and sometimes in lower-speed freight service. Regenerative braking would certainly appear to have been superfluous for pusher service.

But Staufer’s comments on the FF1 were interesting. After noting that the FF1 had two continuous speeds, 10.3 and 20.3 mile/h, he said: “Speed was maintained regardless of grade conditions, drawing power on the upgrade and returning it on the downgrade”. So, the FF1 was regenerative, and it certainly looks as if – back in 1917 - the PRR had envisaged using regenerative locomotives on its planned mountain section electrification. One imagines that as part of its evaluation, the FF1 was able to use regenerative braking on those parts of the PRR system upon which it ran.

Solomon also stated that the late 1930s DD2 prototype was intended for the then-proposed Harrisburg-to-Pittsburgh mountain electrification. It is not immediately obvious why a 1250 hp per axle locomotive with four driving axles would be chosen for mountain service, but if so, then one might have expected some form of electric braking. Alternatively, Staufer describes the DD2 as having been a realization of “universal locomotive” concept, which, with different gear ratios, could be used in both freight and passenger service. In that light, the DD2 looks to have been more like a “bent” version of the R1, with commensurably better tracking and riding.

Cheers,

Pneudyne--
The FF-1 was, I think, the only PRR locomotive ("motor" in PRR-speak) with synchronous motors, and was scrapped early. I don't think any of the later PRR electric locomotive designs allowed for regenerative braking, so I would GUESS that the fixed plant for the main (mid to late 1930s) electrification wasn't set up to deal with regeneration. So my GUESS is that it would have been simpler to isolate the regeneration function on the FF-2 than to modify all the substations, etc, to cope with regenerative braking.
(Note, however, the two capitalized occurrences of "guess" in that paragraph.)

The DD-2 is an interesting locomotive I wish I knew a little more about! My sense is that it was a technology demonstrator or testbed: sized so it could be treated, in operation, as roughly equivalent to a GG-1 or the R-1, but using technology applicable to larger units. I have read SOMEWHERE (I think it was an article in a copy of "Trains" that didn't survive the cull when we cleaned out my mother's house after she died) that the PRR contemplated building a 7500 hp "GG-2" for use either east or west of Harrisburg -- the article I read had an outline drawing as illustration, and I ***think*** I remember it showing, in addition to greater over-all size than a GG-1, larger diameter drivers: 62" instead of 57", and I think I remember from Stauffer that the DD-1 had this size driver -- and even a 10,000 hp 2-D+D-2 that would be used only west of Harrisburg. These notional designs have the same 1,250 hp / driving axle rating as the DD-2, so I have assumed that the DD-2 was a test-bed for them. … Alas, that is all I know. I don't think the article I read had notes about sources or even precise dates. … I'll post another question to the PRR forum: maybe we can get somebody to haul some documentation out of a basement!

Re the PRR FF1, the Wikipedia article (https://en.wikipedia.org/wiki/PRR_FF1" onclick="window.open(this.href);return false;), whilst itself not adding much, includes as attachments a couple of Railway Mechanical Engineer articles from 1917 and 1918 which do provide additional background information.

In brief, it was intended as the prototype for a fleet of mountain locomotives for the planned electrification over the mountains.

The FF1 prototype was of the automatically regenerative type, which anyway would be virtually unavoidable with the split-phase arrangement, and would be the major reason for choosing it.

It was tested on the Philadelphia-Paoli section. Whether its regenerative capability was tested is unclear.

But I think that we may infer that the PRR did plan to use regenerative locomotives on the mountain section had that electrification proceeded, and to achieve that was prepared to accept locomotives that were limited in other ways, at least back in 1917. Projecting on that basis, it might also have tolerated motor-generator locomotives in the 1930s and 1940s. But I think that there is a good chance that it would have looked at developing regenerative locomotives with single-phase motors, following the Swiss example. Then in the mercury arc rectifier era, it might also have leaned upon its suppliers to develop excitron technology to allow regeneration. I suspect though, and perhaps as a hedge, it would also have had locomotives with ignitron rectifiers and dynamic braking, and then silicon rectifiers with dynamic braking would have been something of a forced choice until thyristor technology allowed regenerative braking again.

Against that one may observe that the E2b had dynamic rather than regenerative braking. But it is not clear whether the E2b was designed with electrification of the mountain section in mind. Its quoted “ruling grade” rating of 1060 hp per axle, well above the 625 hp continuous rating, does not appear to be something that could be sustained on long mountain climbs.

Cheers,

What the PRR would have done if they had been able to electrify over the mountains to Pittsburgh is anybody's guess: the prospect of doing so was so far off when the FF-1 was built that it has to be thought of as a "research" locomotive, something whose test results would help with the decision of what to do. Or that would be my guess. (The PRR was willing to put significant amounts of money into experimentation: witness the simultaneous ordering of the R-1 and GG-1 prototypes.)

The much higher (than the continuous rating) short-time rating of the E2b was in line with previous PRR AC "motors". The railroad evidently thought it desirable. (Given their dense traffic in the electrified region, with a mix of through and local passenger, ordinary and mineral freight, I would imagine that they WOULD like something that could overload briefly to accelerate a train back up to full speed after getting a stop signal!) I think they asked GE to build straight AC demonstrators instead of rectifier units like Westinghouse's precisely because they saw value in the operating characteristics of their AC locomotives.

(Sorry, those are sort of obvious comments: hardly a fair repayment for your well-documented facts!)

Sorry, this is something I should have dug out a long time ago!
When the Pennsylvania Railroad's E2b locomotives (for those just joining the discussion now, these are the 1952 experimentals from General Electric: the ones with more than a slight family resemblance to Alco FA diesels) were new, the "Railway Gazette" published a description of them (issue of 18 April, 1952, pp. 432-434).

The article says that the PRR, although they had also ordered rectifier locomotives (not named: these are the Westinghouse E2c and E3b units), felt they needed units with the operating characteristics of their earlier AC "motors," in particular the ability to accept massive overload power on grades and for accelerating trains. (Apparently the E2b could survive brief periods at 1300 hp per traction motor: over twice the continuous rating! But then, at least in folklore, the 4620 hp GG-1 was capable of something like 10,000 hp for brief periods.)

On the topic of this string, the E2b was equipped with dynamic brakes, NOT regenerative brakes: a motor-generator set in the rear of the locomotive provided exciting current, and in braking the traction motors temporarily functioned as DC generators: one suspects that the final braking grids and blowers would have been similar to those GE supplied for Alco diesels.

As to whether the E2b was designed with mountain service in mind... Hope springs eternal in the breasts of those who want railroad electrification, but I think in 1952 it would have been clear to both PRR and GE management that there was no near-term prospect of Harrisbug-Pittsburgh electrification, and the article speaks of renewing the electric locomotive fleet. In the event, much of PRR's electric freight hauling was done by P5a locomotives until they were replaced by GE's E-44 rectifier units in the early 1960s, but my guess would be that PRR, in asking for demonstrators from W'house and GE in the early 1950s, was looking for a possible P5a replacement. (The P5a had, after all, originally been designed for express passenger service, and became a freight locomotive only when replaced by the GG-1 in passenger service. (The GG-1 tracked better, and was less likely to damage itself or the track at high speeds than the P5a.) PRR had originally planned for a fleet of L6 freight electrics, but this was dropped, many partially completed L6 being scrapped. At a guess the L6 was dropped because so many already completed P5a suddenly became available as freight units, but the L6's characteristics suggest that the PRR, in its heart of hearts, thought 625 hp (continuous) per driving axle was more sensible for freight locomotives than the P5a's 1250.)

[attachment=0]PRR E2b RG 1952-04-18 copy.pdf[/attachment]
I'm not sure whether the software of Railroad.net forums allows attachment of 3-page PDF files, but...
Maybe this is the "Railway Gazette" article on the E2b.

The E2b was certainly an interesting locomotive. It may be seen as being broadly similar to typical diesel freight units of the period, with the same B-B wheel arrangement and of about the same weight, but with the advantage of somewhat more continuous power, and a lot more short-term power.

We may only speculate as to whether GE and/or PRR pondered the possibility of equipping the E2b with regenerative braking. Basis Swiss practice of the period, of the available approaches, the exciter motor connection seemed to be the preferred choice where high braking efforts were involved. But that did mean that at least one motor was not available for braking, in turn meaning a reduction in the braking adhesive weight. For example, the SBB Ae6/6, with six parallel connected 16⅔ Hz motors, had five available for regenerative braking, with one used as exciter. But the SNCF CC6051 (Swiss-built), with three parallel groups of series-paired 50 Hz motors, had two pairs available for braking, with one pair used as exciter. The E2b had two parallel groups of series-paired 25 Hz motors, so was perhaps not an ideal starting point for regeneration. A six-motor locomotive would have been better. With single-phase commutator motors, the usable voltage decreases as frequency increases (because of the transformer effect), hence the preference for series pairs at the higher frequencies in order to keep total current draw to manageable levels. It may also be noted that the use of one traction motor as exciter was not unknown in DC regenerative locomotives; it avoided the needed for a separate motor-generator exciter unit that was required to provide relatively high currents, albeit it low voltages.

The E2b motors were fairly big, and evidently required long wheelbase (11’0”) trucks with 48-inch diameter driving wheels to accommodate them. Putting three of them in a C-truck might not have been easy. A triple-truck B-B-B layout might have been easier to implement had a six-motor version been contemplated.

The notion that the E2b was a potential P5a replacement is supported by the fact that the E2b was equipped to operate in MU with the P5a as well as with its own kind. That may explain why it had a 32-volt auxiliary electrical system instead of the 74-volts that might otherwise have been expected. That mixed MU capability probably also predetermined the number of controller steps, namely 21, although that number was commensurate with then-current single-phase motor practice; the Ae6/6 had 28 steps, the CC6051 16 steps. Single-phase motor curves get steeper with higher frequency, thus fewer control steps are required to keep the accelerating current notching increments within given bounds.

Cheers,

The P5a (like most other PRR electric locomotives from the 1930s) had twin-armature motors: each armature was equivalent in power to one of the motors used on the E2b. It would be interesting to compare things like weight between an E2b motor and half of a P5a motor: technology (I suspect the most relevant aspect here would be materials science) had advanced in 15 years, so…

I have received a very interesting e-mail from someone who does not wish to participate in Railroad.net forums and does not wish to be named on them, but who gave me permission to re-post the information he/she provided. Initial greeting and final request for anonymity deleted, the e-mail is as follows:

FF2 regeneration -Trains Magazine June 1958 has a 5 page article by Frederick Westing on the purchase and conversion of these for helper service. On pg 47 he says "But the regenerative braking apparatus was removed, for the work involved made its use unnecessary".

FF1 regeneration -this worked because 3 phase induction motors can act as induction generators simply by applying energy to the shaft and removing it at the electrical terminals. No reversal of rotation nor electrical reconnection is necessary. What is needed however is an existing 3 phase AC supply to push reactive current into the machine to create a rotating "DC" field to excite and generate the returning AC power. The phase splitter neatly does this in the forward direction and converts the 3 phase power to single phase for the catenary in the reverse direction at the same time. Do understand however that if the catenary substation goes dead the locomotive will also die, an inductive generator is not self sustaining without the external supply of magnetizing VARS.

Other AC regeneration -Let us branch a bit here. In broad terms DC machines are equally good as generators and motors. In detail however it is only the shunt field machines that easily generate, series field machines are almost hopeless to control and keep stable. So if we look first at traction (which was almost universally series motors) and then at diesel dynamic brakes; while you already know that the control circuits reconnect the fields in a low voltage DC loop and energize them from the idling diesel generator, the switching to shunt field generation turns out to also be an important step in getting the system to work. This will apply for electric locomotive dynamics on DC roads too, there we will find that the fields are reconnected for shunt generation and then supplied from some separate source. For the CMStP&P EP-3's this was an auxiliary generator mounted on the pilot truck.

Coming forward to locomotives with AC series traction motors however an extra problem occurs as these are commutator machines. If you find and put a DC supply on their field, while they will create AC in the armature, the commutator will chop it to DC which you then have to figure out how to get into the catenary. If you put AC on the field I'm not sure what you would get, I've never seen anybody try to reason it out. I have also read enough about AC series motor design to have grave doubts about managing and commutating the inductive kicks involved too. The issues involve the brushes shorting one coil of the armature as they bridge from one commutator bar to the next while the AC field is trying to power the coil as if it was a transformer winding. The resulting short circuit limits coil design voltages to about 2 volts each. So I think AC regeneration on past series motored locomotives was likely unfeasible, AC commutator motors don't seem to back generate AC.

In the 21st century most of the problems have been eliminated by using 3 phase induction traction motors. With rectifiers/inverters between them and any diesel generator, 3rd rail source or catenary source and any dynamic grid, 3rd rail regeneration, catenary regeneration or HEP load the intermediate DC bus will allow almost any combination that will fit into the carbody.

For more reading I imagine your University access lets you into IEEExplore. Some papers of interest are:
-GN electrification AIEE Transactions Mar 1933
-PRR GE E2B AC series motored locos were built with dynamic braking as you have mentioned. GE claimed this was the first dynamic applied on an electric. The motors were reconnected from series to shunt for braking and an AC-DC MG set excited the fields, so the braking mode functioned in DC. This makes me think GE knew about the difficulties of series motor AC regeneration 60 years before my writing above. AIEE Transactions Jan 1952 pg 27-36
-PRR Westinghouse E3B & E2C DC motored ignitrons were also built with DC dynamic braking. AIEE Transactions Jan 1952 pg 37-47 and Jul 1954.
-NH EP-5/PC E40 DC motored ignitrons did not have electric braking, the New Haven's profile would not have needed it. AIEE Transactions Jul 1955.
-VGN EL-C/NH EF-4/PC E33’s DC motored ignitrons had DC dynamic braking. AIEE Transactions May 1957 pg 68-73.
And elsewhere:
-PRR's FF1 induction motored loco is in The Electric Journal (Westinghouse) vol 14 1917 pg 407-412. Archive.org/details/texts
-PRR E44 DC motored ignitrons had DC dynamic braking (operator manuals). George Elwood's RR-fallenflags.org

Allen, thanks very much to you and your correspondent for that posting, which makes very interesting reading and does answer some of the questions that have come up. I do have some comments and observations, below.

Allen Hazen wrote:FF2 regeneration -Trains Magazine June 1958 has a 5 page article by Frederick Westing on the purchase and conversion of these for helper service. On pg 47 he says "But the regenerative braking apparatus was removed, for the work involved made its use unnecessary".

Good, so we now know the answer to that question about the FF2.

Allen Hazen wrote:Other AC regeneration -Let us branch a bit here. In broad terms DC machines are equally good as generators and motors. In detail however it is only the shunt field machines that easily generate, series field machines are almost hopeless to control and keep stable. So if we look first at traction (which was almost universally series motors) and then at diesel dynamic brakes; while you already know that the control circuits reconnect the fields in a low voltage DC loop and energize them from the idling diesel generator, the switching to shunt field generation turns out to also be an important step in getting the system to work. This will apply for electric locomotive dynamics on DC roads too, there we will find that the fields are reconnected for shunt generation and then supplied from some separate source. For the CMStP&P EP-3's this was an auxiliary generator mounted on the pilot truck.

Yes, I’d say that that shunt (separately excited) machines were the primary type used for dynamic brakes, and very probably the only type used for diesel-electric locomotive dynamic brakes, for DC electric locomotive regenerative brakes, and for AC-DC electric locomotive dynamic brakes. But perhaps not to be overlooked is that self-excited series machines were used for dynamic braking in streetcars (including the PCC), heavy rail transit equipment, and in some DC electric locomotives. So the evidence is that series excitation could in fact be made to work stably enough for braking purposes, albeit within boundaries.

Allen Hazen wrote:Coming forward to locomotives with AC series traction motors however an extra problem occurs as these are commutator machines. If you find and put a DC supply on their field, while they will create AC in the armature, the commutator will chop it to DC which you then have to figure out how to get into the catenary. If you put AC on the field I'm not sure what you would get, I've never seen anybody try to reason it out. I have also read enough about AC series motor design to have grave doubts about managing and commutating the inductive kicks involved too. The issues involve the brushes shorting one coil of the armature as they bridge from one commutator bar to the next while the AC field is trying to power the coil as if it was a transformer winding. The resulting short circuit limits coil design voltages to about 2 volts each. So I think AC regeneration on past series motored locomotives was likely unfeasible, AC commutator motors don't seem to back generate AC.

Well, that’s what I thought too until I read about single-phase commutator motor regenerative locomotives being used in Switzerland. Even then I initially saw it as an easily dismissed oddity, whereas in fact it was an established practice going back to the early days of single-phase motors, with initial development work being done by Behn Eschenburg, who was a major player involved in bringing the low-frequency single-phase motor to a workable form for traction purposes. There certainly were difficulties as compared with DC regeneration practice, but they were overcome by various means. And regenerative braking was used by the SBB right through until its final design of single-phase commutator locomotive, the Re6/6, first built in the 1970s, with construction continuing into the 1980s. I’ll follow up with some book excerpts that provide an overview of single-phase regeneration techniques.

Allen Hazen wrote:For more reading I imagine your University access lets you into IEEExplore. Some papers of interest are:

-PRR GE E2B AC series motored locos were built with dynamic braking as you have mentioned. GE claimed this was the first dynamic applied on an electric. The motors were reconnected from series to shunt for braking and an AC-DC MG set excited the fields, so the braking mode functioned in DC. This makes me think GE knew about the difficulties of series motor AC regeneration 60 years before my writing above. AIEE Transactions Jan 1952 pg 27-36

Certainly this was the first dynamic brake applied to a North American AC electric locomotive, and possibly to any North American electric locomotive; I can’t think of any DC examples that had dynamic rather than regenerative braking. I imagine that GE was referring to North American practice in claiming a “first”. Worldwide, dynamic braking (usually referred to as rheostatic braking in Europe when used on electric locomotives) was well-established going back to the 1920s at least, and for both DC and AC (single-phase commutator motors) locomotives. In the AC case, both AC and DC excitation were used. I imagine that one advantage of AC excitation was that it avoided the need for a heavy current DC machine. In Switzerland, whereas SBB favoured regeneration, BLS used dynamic braking on its single-phase electric locomotives, with AC excitation on some of the older types but DC excitation for the Ae4/4 and later. From the apparent SBB and BLS dichotomy, one might reasonably infer that the choice between regenerative and dynamic braking was quite situational, with no clear winner in the abstract. Possibly then GE’s choice for the E2b was more situation-based than feasibility-based. I am not sure that it did, but if the E2b had employed standard diesel-electric locomotive dynamic braking units, that could have been another reason for GE’s choice.

Cheers,

Here are some pages from the Dover book (details following) that outline some of the problems and their solutions in respect of regeneration with AC single-phase motors.
Cheers,

And here are some pages from the Brooks book that although somewhat overlapping Dover, provide a broader view of electric braking with AC single-phase motors.
Cheers,

Here are the details for the Brooks book, plus some more excerpts dealing with DC dynamic braking with series-connected motors.
Cheers,