3rd rail vs OHLE

[O L Leigh]

The new kid on the block is not a “consultation” as mentioned above by O.L, although that may follow.

The possible game changer may be a consultant’s report into the economics of making 3rd rail safer.

Network Rail who have advertised for consultants to carry out the study made their position clear by saying at the start of the advert that third rail is ten times more dangerous.

That is the consultation to which I was making reference, the advertisement for which kicked off the discussion last year.

Was this retraction published, perchance?

I've seen no retraction so would welcome a lead on this.

Interesting when did they advertise for this service?

It was June of last year. Click.

I'll merely point out that the 1950s battery twin set operated on Aberdeen to Ballater, charging overnight in the bay at Aberdeen with a bit of a boost when turning round there, was 43 miles each way, 86 miles round trip, or 139km. It did three or four round trips a day, to the same timings as the dmu which operated the opposite diagram. The battery industry constantly goes on about how they are advancing, but really things are not that far forward in 60 years.

I don't think there was much on that unit to draw power from the battery, though. The motors, obviously, a few light bulbs, presumably a vacuum pump for the braking system and that would have been about it. Not much more complex than a contemporary milk float, really.

You are correct in all this. The Kentish Town incident a few years ago showed that, for all the modern emergency batteries were rated to a certain spec, on the day they delivered less than half that.

This is quite a hobby-horse of yours, isn't it.

Without wishing to invite further topic drift, the batteries on the units in question were not intended to run hotel power such as A/C, still less move the train, and so were not rated to give that sort of output. A/C is a very high-draw system and the provision of battery power sufficient to keep a packed train cool on a hot day for even a couple of hours off the juice would require quite some additional capacity with it's consequence for cost and weight. And all just to cover off an eventuality that an individual unit may never have to cope with but which would penalise the operator every single day in terms of additional maintenance cost, track access charges and so on.

 

[Taunton]

They were lead acid cells but added 16T to train weight and they had 60min+ layovers either end for limited charging mind you although takes a fair time to recharge lead acid cells but this was pretty impressive given the basic technology and no regen. Guess it was go anywhere diesels that seemed much more straightforward at that time as to why it didn't catch on unlike the German Class 515's which ran into the 100's and lasted til early 90's.

Alas the sands of time cause the detail to be ever expanded ...

Turnround time at Ballater (1961 timetable attached) :

Morning run - 26 mins
Afternoon run - 10 mins
Evening run - 33 mins

Hardly time to get the cables on and off again in the afternoon. So I guess they didn't.

I always thought it was pretty impressive as well for the era. Those German battery units (both DB and the GDR) indeed ran fine as well.

 

[Bald Rick]

You have to allow for hotel load which on a hot day will need 30-40kwh/coach to run air con etc

Yep that was in the modelling.

then you need to allow for the potential for sitting down somewhere due to an operating incident so you need to have a margin in the "tank" as well

Some margin yes, and that is provided, even if the Uckfield charging station is offline. However yo7 don’t need that much. If a train is ‘stranded’ for a long time on the way out, it would, as now, turn back short and get back North. In an absolute worst case where one is stranded for hours, firstly the passengers would need detraining within an hour or so so the unit can be largely shut down. Secondly with that sort of disruption the service would be suspended, so if a unit was stranded and didn’t have enough (or for some reason, any) power to get back to Hurst Green, a rescue unit would be deployed to meet it.

 

[mike57]

You are correct in all this. The Kentish Town incident a few years ago showed that, for all the modern emergency batteries were rated to a certain spec, on the day they delivered less than half that.

Its not just the rail industry that suffers from the 'rose tinted glasses' effect when considering battery systems. The calculations I saw and challenged on a couple of occasions had insufficent margins for error. Optimum charge regime, new batteries, no high or low temperatures considered to list a few. In both cases the battery capacity ended up being nearly doubled, and even then the reality was that after a few years the system would only just maintain its required back-up time. Both cases were 10's of kW back up systems, and in one case we added a diesel generator, and retained the batteries to give time to start the generator. When designing systems for the real world you need to consider whats the worst that can happen, to be fair in the cases I worked on the consequences of failure could be significant, but I suspect rolling stock traction/power system would fall into the high consequence category as well.

 

[norbitonflyer]

One unit remains and is still used at the Deesside railway,

There only ever was one unit.

 

[contrex]

DB had done some AC OHL before the war, at 20kv rather than 25kv

DR I think?

 

[a_c_skinner]

The reality is the number of 'in-fill' schemes is actually very small:

Ashford - Ore
Hurst Green - Uckfield

Perhaps the North Downs line ?

Isn't this the key point? There are fewer than a handful of places where the gain of 3rd rail is big, in a part of the world where it is widely in use and these seem to the casual crayon user pretty obvious choices.

 

[rower40]

Time for some anecdata...
In the late 1980s or early 1990s, I was waiting one Sunday morning at Beaulieu Road station in deepest (stops to google) Hampshire.
A track maintenance gang arrived in a van, then proceeded to cross the line at the foot of the platform ramp. One of the team tripped over his bootlace, and I'll never forget the look of sheer terror on his face as he fell towards the powered third rail. He managed to get his foot out to stop his fall, so no harm done. He then got to a position of safety (the foot of the other platform ramp) and sat down to secure his bootlace, shaking all the while.

 

[contrex]

Am I right to think that new third rail installations are banned, except for short extensions/new replacements?

Overall OLE is more efficient, as well as not having to worry about power supply constraints*

*Unless you send a 92 down a suburban commuter line

I fondly remember one of my first Eurostar journeys from Waterloo to Paris around 1998. As we ran alongside the South London Line I saw a freight train hauled by a class 92. We didn't get past it until Brixton.

 

[Annetts key]

Its not just the rail industry that suffers from the 'rose tinted glasses' effect when considering battery systems. The calculations I saw and challenged on a couple of occasions had insufficent margins for error. Optimum charge regime, new batteries, no high or low temperatures considered to list a few. In both cases the battery capacity ended up being nearly doubled, and even then the reality was that after a few years the system would only just maintain its required back-up time. Both cases were 10's of kW back up systems, and in one case we added a diesel generator, and retained the batteries to give time to start the generator. When designing systems for the real world you need to consider whats the worst that can happen, to be fair in the cases I worked on the consequences of failure could be significant, but I suspect rolling stock traction/power system would fall into the high consequence category as well.

I don’t doubt that there have been some poorly designed systems. But there will be plenty of battery systems that have met the requirements made of them and hence have been successful. And because they have worked without problems, you don’t normally hear about them. True, mostly on a smaller scale than a system designed for a trains propulsion system.
The maximum charging rate, the temperature performance range, and the rate of loss of capacity over time very much depends on the technology of the cells used. The various different battery technologies having significantly different characteristics. All of which should be carefully considered at the system design stage. The quality of the cells is obviously very important as well.

So although different technology, this is no different to the principles that apply to other power systems, such as the problem of weight and range of early aircraft using ICE and then later on, when first using jet engines. We learned from the prototypes, continued to learn from the first production units and improved the technology. Now we are very good at managing the remaining limitations of the technology. Why would the same not apply to battery powered trains?

 

[AM9]

... I don't see charging on the move too much of an issue as the train will be able to divert power to charging when its not motoring and reasonably guaranteed load when its regening. Current schedule gives around 75mins on the juice so ought to half that time with full load available for charging. ...

That sounds OK but consider the problem at system level, where the electrified part of the route can be pretty heavily loaded with conventional DC trains and is rated assuming diversity of those trains' demand, i.e. their non-motoring has already been used to calculate the total system power the installed equipment needs to provide. Load that supply up with a few battery charging trains with their near continuous demand could well require the total route needing uprating.

 

[Nicholas Lewis]

That sounds OK but consider the problem at system level, where the electrified part of the route can be pretty heavily loaded with conventional DC trains and is rated assuming diversity of those trains' demand, i.e. their non-motoring has already been used to calculate the total system power the installed equipment needs to provide. Load that supply up with a few battery charging trains with their near continuous demand could well require the total route needing uprating.

Fair point but with an hourly Uckfield line service a couple of four cars will be noise on the grid points that service London Bridge to Hurst Green. However, wider adoption could cause issues as you could see ORR saying oh now you have battery power you can de-electrify lines so the overall risk of operating the third rail system falls.

That is the consultation to which I was making reference, the advertisement for which kicked off the discussion last year.



I've seen no retraction so would welcome a lead on this.



It was June of last year. Click.

Thanks for link it was RSSB calling for consultancy support and the remit is pretty comprehensive to cover off all the issues and come to some conclusion once and for all. Report not due till 2021.

 

[GLC]

That sounds OK but consider the problem at system level, where the electrified part of the route can be pretty heavily loaded with conventional DC trains and is rated assuming diversity of those trains' demand, i.e. their non-motoring has already been used to calculate the total system power the installed equipment needs to provide. Load that supply up with a few battery charging trains with their near continuous demand could well require the total route needing uprating.

In a similar vein however, with admittedly careful planning (and corresponding logic from the train), a BEMU could draw on its battery reserves rather than using up 3rd rail supply in congested areas, and postpone charging until a time/area which is under less stress. Obviously this would have to be carefully balanced to make sure there is enough energy to complete the unelectrified parts of a diagram, but the possibility is there

 

[AM9]

In a similar vein however, with admittedly careful planning (and corresponding logic from the train), a BEMU could draw on its battery reserves rather than using up 3rd rail supply in congested areas, and postpone charging until a time/area which is under less stress. Obviously this would have to be carefully balanced to make sure there is enough energy to complete the unelectrified parts of a diagram, but the possibility is there

Still not that easy, The line to South Croydon is pretty tied up in terms of power capacity, the line from there south to Hurst Green is power rated just for the East Grinstead service, - in the peaks, two trains per hour. Presumably served by Electrostars but how many cars per train in the peak? The Uckfield line runs 10 or 12 car trains in the peak so with increasing power demand from the BML services, the Uckfield trains might have to get up to 50 miles worth of charge in the 10 miles between South Croydon and Hurst Green, - sharing a supply that has to date be sized for the East Grinstead services alone.

 

[Bald Rick]

Fair point but with an hourly Uckfield line service a couple of four cars will be noise on the grid points that service London Bridge to Hurst Green.

Hurst Green to S Croydon is an issue. It would be an issue if Uckfield was electrified conventionally of course. You might remember the night in 2006 when the whole branch tripped due to overload on a hot night when the World Cup was on and the local power network was under strain. It may even have been 15 years ago today (when we got knocked out by Portugal on penalties).

However, wider adoption could cause issues as you could see ORR saying oh now you have battery power you can de-electrify lines so the overall risk of operating the third rail system falls.

I think that is very unlikely in the near term.

 

[Nicholas Lewis]

Still not that easy, The line to South Croydon is pretty tied up in terms of power capacity, the line from there south to Hurst Green is power rated just for the East Grinstead service, - in the peaks, two trains per hour. Presumably served by Electrostars but how many cars per train in the peak? The Uckfield line runs 10 or 12 car trains in the peak so with increasing power demand from the BML services, the Uckfield trains might have to get up to 50 miles worth of charge in the 10 miles between South Croydon and Hurst Green, - sharing a supply that has to date be sized for the East Grinstead services alone.

The E.Grinstead line had a power enhancement to support Class 700's and 12 car Electrostars so trackside substation capacity is there to support train lengths. Furthermore, whether a train is drawing power for traction or charging won't impact trackside substations. What will be impacted is whether the distribution system between Sth Croydon and E.Grinstead (Dormansland is grid connection) can support the additional load the Uckfield service would impose - it certainly wouldn't have done in 1986 as we designed for 8 car EP stock on a half hourly service but parts of it have been connected to the 33kv system since.

The other thing to consider on charging is the train doesn't have to fast charge when its on the juice as it has 75mins round trip to Hurst Green so charging rate could be massively reduced so it would appear as noise in amongst all those 12 car 700's.

 

[hwl]

The E.Grinstead line had a power enhancement to support Class 700's and 12 car Electrostars so trackside substation capacity is there to support train lengths. Furthermore, whether a train is drawing power for traction or charging won't impact trackside substations. What will be impacted is whether the distribution system between Sth Croydon and E.Grinstead (Dormansland is grid connection) can support the additional load the Uckfield service would impose - it certainly wouldn't have done in 1986 as we designed for 8 car EP stock on a half hourly service but parts of it have been connected to the 33kv system since.

The other thing to consider on charging is the train doesn't have to fast charge when its on the juice as it has 75mins round trip to Hurst Green so charging rate could be massively reduced so it would appear as noise in amongst all those 12 car 700's.

As previously covered several times recently, the upgrade for 700s was the minimum possible and the supply etc. south of Croydon is maxed out again post upgrade. Uckfield 3rd rail would need another upgrade north of Hurst Green and Battery an even bigger upgrade north of Hurst Green.

 

[Taunton]

I don't see how there is much difference between the 33kV lineside power feeder, and doing it at 25kV through the overhead. And the periodic DC lineside substations just match a number of what is under the floor of a 25kV emu. I'm not an electrical engineer, so please point out the difference if there is any.

I do see that a hybrid battery/emu is going to draw double the current (OK, percentages might be wrong) of a plain emu once it gets onto the third rail, because it will be drawing for propulsion plus battery recharge.

 

[Aictos]

If I recall correctly, voltage changeover on a Cl313 was simply a matter of dropping/raising the pan. If there was AC with the pan raised the unit would run on that, otherwise it would take DC through the shoes. Therefore there would be no way to check if the changeover would be successful unless the train was in contact with both systems. The first indication the driver would get of a problem would be a loss of the line light. I would imagine that modern trains would be broadly similar, except that the TMS may indicate a problem with raising the pan due to an ADD operation.

Just to add to the post above to explain that the driver would get a buzzer in the cab heading to Moorgate as the train would be in contact with both the OHL and the 3rd Rail, this also would apply at Drayton Park from when you selected AC as the 3rd rail would still be in contract with the train.

Remember the 3rd rail shoes on a Class 313 are non retractable like the Class 700s which is why they were restricted from going north of both Biggleswade or Royston until infrastructure works was completed to allow Class 700s to work north of those stations.

As to check if the changeover was successful or not, there was two things that happened that you could see and hear, one was the lights in the passenger saloon would go out then back on again and the motors/compressors would change their sound so it was possible to know if the changeover was successful or not.

 

[AM9]

I don't see how there is much difference between the 33kV lineside power feeder, and doing it at 25kV through the overhead. And the periodic DC lineside substations just match a number of what is under the floor of a 25kV emu. I'm not an electrical engineer, so please point out the difference if there is any.

I do see that a hybrid battery/emu is going to draw double the current (OK, percentages might be wrong) of a plain emu once it gets onto the third rail, because it will be drawing for propulsion plus battery recharge.

Unless the 33kV feeder is cooled, it will have to be far greater cross section than an air cooled OLE conductor wire, then there's the additional losses from low voltage track electrification vs high voltage OLE.*

*Technically 25kV (and 33kV) is medium voltage but the argument is the same.

 

[mike57]

Also remember that with larger conductors on the 33kV feed the skin effect becomes significant, so increasing the size of the conductor does not give the expected drop in resistance, once the conductors exceed about 20mm dia.

 

[AM9]

Also remember that with larger conductors on the 33kV feed the skin effect becomes significant, so increasing the size of the conductor does not give the expected drop in resistance, once the conductors exceed about 20mm dia.

Agreed, 33kV is not really suitable for delivering power with multi megawatt short term demands over distances of 10+miles. OLE has multiple smaller conductors of that size inherent in its design (i.e. although not all ideal materials, there would be two continuous paths per track, - conductor and catenary with frequent droppers to distribute the difference).

 

[hwl]

Also remember that with larger conductors on the 33kV feed the skin effect becomes significant, so increasing the size of the conductor does not give the expected drop in resistance, once the conductors exceed about 20mm dia.

Which is why multi-strand cables are used are for the standard 300mm^2 CSA cables for NR.
The only single core 33kV is lower CSA and has aluminium conductors.

 

[Nicholas Lewis]

Agreed, 33kV is not really suitable for delivering power with multi megawatt short term demands over distances of 10+miles. OLE has multiple smaller conductors of that size inherent in its design (i.e. although not all ideal materials, there would be two continuous paths per track, - conductor and catenary with frequent droppers to distribute the difference).

A three phase 33kV power system can deliver a lot more electrical power than a single phase 25kV overhead line (this is over one track) over a longer distance. The 33kV system supplying the Southerns trackside traction substations through a trackside cable network radiates out from grid points. Nominally the cables are 185sqmm Al so can support 50MW compared to 15MW/single OHL. The limiting factor on how you supply a trackside substation from 33kV is the voltage drop but under outage conditions a grid feeder can easily reach 40miles on lines with 2-3 TPH but would be less down the London Bridge to Three Bridges route due to much higher traffic density.

If your talking about delivering power to individual trains there's no dispute that 25kV wins out comprehensively.

 

[apk55]

With a railway electrification system one of the limiting factors is return voltage drop. If this exceeds a few volts then you run a real heath and safety issue of electric shocks and stray currents. The much lower currents with high voltage AC systems make it less of a problem and there are techniques available such as booster transformers or +-25KV systems that can further reduce the problem, On tramways for example they often bury a wire mesh beneath the tracks to intercept stray currents and often in critical locations have very close substation spacing.

 

[Pigeon]

IIRC you can expect a hundred or maybe even two hundred volts on the running rails under worst case conditions with the Southern 750V system. For making any serious attempt to bring the resistive losses down to a level comparable with 25kV systems, the return loss is as important as the feed loss. It can be done, using large section aluminium conductors and a somewhat higher voltage, but without making the voltage increase unfeasibly large it does necessitate fourth rail return.

This is not so terrible, though, because it means you automatically end up with the advantages regarding stray ground currents and interaction with signalling circuitry that they chose the four rail system for the Underground to get. It's a solve one problem, get two sorted free kind of situation.

 

[Nicholas Lewis]

IIRC you can expect a hundred or maybe even two hundred volts on the running rails under worst case conditions with the Southern 750V system. For making any serious attempt to bring the resistive losses down to a level comparable with 25kV systems, the return loss is as important as the feed loss. It can be done, using large section aluminium conductors and a somewhat higher voltage, but without making the voltage increase unfeasibly large it does necessitate fourth rail return.

This is not so terrible, though, because it means you automatically end up with the advantages regarding stray ground currents and interaction with signalling circuitry that they chose the four rail system for the Underground to get. It's a solve one problem, get two sorted free kind of situation.

Your not permitted voltages at that level due to touch potential issues with people touching the train bodywork so return bonding needs to be suitable designed. You do experience approaching 100V at somewhere like Seaford as its remotely fed down a single line but on a two track railway it will be of the order of 50V as you have four rails. Driving the voltage up to say 850V won't help the losses thats the beauty of 25kV your containing all the high currents within the traction unit.

 

[AM9]

Your not permitted voltages at that level due to touch potential issues with people touching the train bodywork so return bonding needs to be suitable designed. You do experience approaching 100V at somewhere like Seaford as its remotely fed down a single line but on a two track railway it will be of the order of 50V as you have four rails. Driving the voltage up to say 850V won't help the losses thats the beauty of 25kV your containing all the high currents within the traction unit.

Wow! I've always known that return path voltage drop is a problem, but a conducting surface at a potential of 50V, is above that allowed for an external earthed eguipment. So what happens when a train pulls into a station and passenger touches the car body when it pulls into a station?

 

[grumpyoldman01]

And the Great Western also toyed with electrification west of Taunton in 1938 (presumably also at 1,500 DC overhead) to eliminate steam working - both to avoid the cost of coal transport, and because of the operating challenges of the Wellington and South Devon banks.

The GWR planned to use 3000v DC overhead; there's quite a bit about it in Oswald Nock's 'The Great Western Railway in the 20th century' (pages 153 et al)

They planned to electrify the main line from Taunton through to Penzance, together with the Torwuay, Kingswear and Brixham branches, Par to Newquay, and the loop from Lostwithiel through Fowey to Par; Merz & McLellan were their consulting engineers for the project

The reality is the number of 'in-fill' schemes is actually very small:

Ashford - Ore
Hurst Green - Uckfield

Perhaps the North Downs line ?

I'm not sure Merseyrail justifies extension - particularly the ones which keep coming up i.e. Wigan - Kirkby and Ormskirk - Preston.

In some ways it's a pity that the Uckfield line can't take Mk3 EMUs because 769s would actually work quite nicely there - you could run them on 3rd rail to Hurst Green then diesel to Uckfield.

Wouldn't Ashford - Ore be better at 25kv AC as there are aspirations to operate services over it to St Pancras via HS1?

Wigan - Kirkby was initiated and traction equipment for class 509s for it was actually ordered, but the plug got pulled on the project - I never learned the reason why

I've seen it stated elsewhere that the Uckfield line can't take Mk 3 EMUs; does anybody know the reasons why? Are the problems preventing them from running major?

 

[Bald Rick]

So what happens when a train pulls into a station and passenger touches the car body when it pulls into a station?

There was an issue at Blackfriars before the humps were introduced with the wheelchair ramps. The insulation wore off the underside of one, and a member of staff put one onto a train with the bottom of the ramp touching a metal access cover. I think it effectively welded itself.