Converting 3rd Rail infrastructure to 25kv AC OHL infrastructure

[HSTEd]

In the case of the Uckfield branch, he said the exact opposite (p61 of the report):

On page 55 is that there is no significant difference between the costs of the AC and DC scheme, he then gives largely unsupported suppositions that AC is cheaper than DC.

Separate experts have suggested to me that a cost in the range of £75m-£95m is more likely for electrifying Uckfield – Hurst Green alone, whether this is with AC overhead or DC third rail

(In a Modern Railways interview, he expanded on the substations point: Substations would also require land to be purchased to put them on).

Meanwhile AC systems would be fed with a magical supergrid bulk supply point that consumes no land at all

Indeed in the report he mentions that it is entirely possible to electrify using DC without a national grid connection at all, instead drawing power from the local distribution network at 11kV or 33kV.

The line between Sanderstead and East Grinstead was electrified as cheaply as possible in 1987, and the 11kV power supply

Tram-style 750Vdc rectifier plants are extremely compact.

A 92 can only deliver 4MW if that power is available in the third rail. The consequence of using 750V DC is that it's much harder to provide that sort of power efficiently. Megawatt-level loads really do take 750V DC third rail well beyond its comfort zone, while 25kV wouldn't be breaking a sweat. Of course, you can mitigate this by providing lots of feeder stations, but the train interface is still never going to work well.

Yes, but you can buy an awful lot of substations for the cost of ripping it all out and providing a whole new grid supply system, distribution system, all the steelwork and resignalling works, and all the associated disruption.

EDIT:
Apologies, I got the Gibb Report and the Uckfield Line mixed up with the Kent Route Study - Technical Appendix draft and the Marshlink LIne.

Page 20/58:

6.2.6. The options are expected to cost:
• Option 1 - dc electrification £100-250M
• Option 2- ac electrification £250-500M
• Option 3 - no direct cost.

 

[Mikey C]

The first part - not so. Both systems were live on the two track section between Channelsea Jn and Cannonbury for about 25 years until the line was closed for a few months for upgrading in 2010/11. The 313s usually ran on AC, changing at Hackney Wick and Dalston. However if there was ever a problem with the AC, they would stay on DC. The track circuits down there were a particular design and frequency to cope, however it was very much preferred for trains to be in AC to limit return current problems. Even so, with 3-4 X 3 car trains an hour, the return current issue was nothing as significant as on Thameslink.

The corrosion issues were worst at Acton, where the AC DC changeover was in the same area as the voltage gradient (where NR DC went from 660 - 750v), lots of other stray currents from the GWML, WCML, LUL, etc etc. This bit had al the masts replaced in 2004 - I poured the concrete on some of the new foundations!

The reason for conversion of Cannonbury - Channelsea was mostly because the anticipated increase in service for London Overground: double the frequency, each 378 roughly double the power of the 313s, meant that the DC system did not have sufficient capacity. So it was taken out, and eliminated three voltage changeovers at the same time (Hackney Wick, Dalston, Camden)

When the East London Line was extended to Dalston it was done via 3rd rail which continues to Highbury & Islington, I'm guessing this must be the last 3rd rail installation, I assume it was done via 3rd rail for operational convenience, avoiding having to chance currents at Shoreditch High Street?

 

[rebmcr]

I'm guessing this must be the last 3rd rail installation

Soon to be Battersea on the Northern Line Extension.

 

[Bald Rick]

Soon to be Battersea on the Northern Line Extension.

Which is 4th rail

 

[rebmcr]

Which is 4th rail

Where there is a fourth, there must also be a third!

Realistically though, the presence of a rail-level conductor is all that really matters here.

 

[edwin_m]

The projected Bakerloo Line extension will presumably be fourth rail.

 

[Taunton]

There was a study about five years ago that claimed it was no more expensive to convert to 25kV than to renew a lot of the SWML 750V infrastructure that was approaching end of life

Sorry, but that report (which I seem to recall) was patent nonsense. I presume the DfT saw it the same way.

Was it done by the same estimators who did the GMWL costings?

(In a Modern Railways interview, he expanded on the substations point: Substations would also require land to be purchased to put them on).

I think the whole set of substations on the traditional Southern lines were sited on existing railway land. Why would any new electrification be different?

It's well known that what once required many tons of equipment in a 1960s 25Kv loco for rectification, transformer and control can now be done on a simple raft. How is it that substations, which essentially do the same thing, have not been commensurately improved?

Incidentally, 25Kv does not do away with the need for substations. The power station supply is still at high voltage and the train motors still work at approx. 750v. It's just with 25Kv a 12-car train drags round three substations with it all day long.

 

[rebmcr]

It's just with 25Kv a 12-car train drags round three substations with it all day long.

And a 750v 12-car train carries around the same weight in concrete for the purposes of harmonising suspension settings across fleets.

 

[SpacePhoenix]

Sorry, but that report (which I seem to recall) was patent nonsense. I presume the DfT saw it the same way.

Was it done by the same estimators who did the GMWL costings?

I think the whole set of substations on the traditional Southern lines were sited on existing railway land. Why would any new electrification be different?

It's well known that what once required many tons of equipment in a 1960s 25Kv loco for rectification, transformer and control can now be done on a simple raft. How is it that substations, which essentially do the same thing, have not been commensurately improved?

Incidentally, 25Kv does not do away with the need for substations. The power station supply is still at high voltage and the train motors still work at approx. 750v. It's just with 25Kv a 12-car train drags round three substations with it all day long.

And a 750v 12-car train carries around the same weight in concrete for the purposes of harmonising suspension settings across fleets.

dc if the substation fails, all dc units/locos supplied by it will be stuck (unless the section can be fed by another). With ac if the "substation"/transformer on the unit or loco fails, only that unit or loco will definitely be stuck, all other ac units or locoas can still move, unless they're trapped by the failed unit or loco

 

[The Ham]

Sorry, but that report (which I seem to recall) was patent nonsense. I presume the DfT saw it the same way.

Was it done by the same estimators who did the GMWL costings?

I would generally assume that anyone doing a comparison of costs would have costed both projects. As such there is a risk that they under estimated the costs for the third rail replacement as well.

There is also an argument that to directly replace third rail you have to have it up and running at the end of the line closure, which means if you got a problem it could lead to major delays. Whilst if you are building a new power supply system you just walk away at the end of closure and sort out the next day/week.

This means that it could be done much faster (as you can more easily have overnight works) and therefore you reduce the costs of rail replacement buses.

 

[HSTEd]

dc if the substation fails, all dc units/locos supplied by it will be stuck (unless the section can be fed by another). With ac if the "substation"/transformer on the unit or loco fails, only that unit or loco will definitely be stuck, all other ac units or locoas can still move, unless they're trapped by the failed unit or loco

So in almost all cases the DC units will be able to move

There is also an argument that to directly replace third rail you have to have it up and running at the end of the line closure, which means if you got a problem it could lead to major delays. Whilst if you are building a new power supply system you just walk away at the end of closure and sort out the next day/week.

Most of the work on a DC system is entirely off the track - you can just build the new substation next to the old one.
The actual unplugging of the old substation and plugging in of the new one is a tiny fraction of the work.

And the track work is regularly performed as part of regular track maintenance - it is far less susceptible to delays.

This means that it could be done much faster (as you can more easily have overnight works) and therefore you reduce the costs of rail replacement buses.

Except replacing a substation with a newbuild adjacent substation does not interrupt the operation of the railway at all.

 

[JohnElliott]

I think the whole set of substations on the traditional Southern lines were sited on existing railway land. Why would any new electrification be different?

Without delving in my recycling bin for the magazine (if it's even still in there) I can't be certain, but I think it's because the Uckfield line was built relatively late (1880s or 1890s) and on a low budget, so there isn't that much spare railway-owned land around the line.

Incidentally, a comment on London Reconnections gives another expense that would apply to DC electrification to Uckfield: the track circuits are DC, so would be immune to AC but not to DC. Obviously that's specific to the Uckfield branch; for a DC->AC conversion rewiring the track circuits would be another expense on the AC side of the ledger.

 

[JohnElliott]

dc if the substation fails, all dc units/locos supplied by it will be stuck (unless the section can be fed by another).

The 2-3 mile spacing of substations (per the Gibb report) is partly so that if one fails, the two either side can still provide enough juice to keep the trains moving. When the Haywards Heath - Horsted Keynes branch was electrified, it had no substations so trains departing Horsted Keynes would have been drawing power from at least 5 miles away.

 

[edwin_m]

Sorry, but that report (which I seem to recall) was patent nonsense. I presume the DfT saw it the same way.

With the way 25kV scheme costs have gone I'd be inclined to agree with you. However I do live in hope that they will be brought under control at some point.

 

[Mikey C]

Where there is a fourth, there must also be a third!

Realistically though, the presence of a rail-level conductor is all that really matters here.

I was ignoring Underground extensions, where there's no choice (the Croxley scheme would be another one if it ever happens!)

In the case of the East London Line extension, this could have been OHLE, as it effectively connects a 3rd rail line with an OHLE line.

 

[Domh245]

In the case of the East London Line extension, this could have been OHLE, as it effectively connects a 3rd rail line with an OHLE line.

You'd have run into clearance issues with things like the Thames tunnel, probably even most of the route as it'll have only been tunneled to a certain size that wouldn't have allowed for overhead equipments

 

[InOban]

Brunel's Thames tunnel is over 6m high. Would that not have accommodated OHLE?

 

[Mikey C]

You'd have run into clearance issues with things like the Thames tunnel, probably even most of the route as it'll have only been tunneled to a certain size that wouldn't have allowed for overhead equipments

I was thinking more of the surface section, the new part north of Whitechapel which is effectively new railway

 

[QueensCurve]

I suppose that the 25kV clearance requirements may come under some review if the hardline EU leavers get their way.

Leaving the EU is a red-herring.

The European standards recognise the difficulties, not to say impossibility, of providing standard clearances on the British Railway. They problem is that the UK Regulators chose to copy and paste the new European standards rather than use the freedom that they had.

There is ample scope for clearances to be reduced even within the existing standards by use of risk assessment. There are some reported signs that the UK regulators are open to reconsideration.

While the unthinking implementation of European standards is undoubtedly one factor in the present crisis in electrification, there are also the questions of overspecification of the equipment and the competence in delivering the project.

 

[swt_passenger]

And a 750v 12-car train carries around the same weight in concrete for the purposes of harmonising suspension settings across fleets.

Not every fleet. DC Desiros are not ballasted.

 

[QueensCurve]

I saw the designs for the BML in the presentation!

Just for you I went and managed to track down a copy still online. Page 15! https://www.imeche.org/docs/default...to_AC_Conversion_by_Steve_Yianni.pdf?sfvrsn=0

Many thanks for posting this. It is a vital resource.

 

[swt_passenger]

dc if the substation fails, all dc units/locos supplied by it will be stuck (unless the section can be fed by another)...

The "DC system" does normally allow for many adjacent sections connected in parallel. That's what the equipment in the 'track paralleling huts' does.

 

[Taunton]

The European standards recognise the difficulties, not to say impossibility, of providing standard clearances on the British Railway. They problem is that the UK Regulators chose to copy and paste the new European standards rather than use the freedom that they had.

There is ample scope for clearances to be reduced even within the existing standards by use of risk assessment. There are some reported signs that the UK regulators are open to reconsideration.

There is some disappointing self-serving justification in a letter in this month's Modern Railways from the head of one of the regulation bodies (there seem to be several overlapping ones and it is a bit difficult to keep up) who seems to think they have done everything within their powers perfectly.

 

[swt_passenger]

I was thinking more of the surface section, the new part north of Whitechapel which is effectively new railway

It's the low proportion of new track (starting between Whitchapel and Shoreditch) compared with the four branches on the "Southern" - including the Clapham Jn route.
DC was the obvious choice back when the ELL extension was designed, because the "presumption against third rail" hadn't happened yet.

 

[HSTEd]

It appears the ORR, or it's antecedents, should never have been created at privatisation.

The entire safety regulation and standardisation job should have been turned over to HMRI as it existed at that time, they at least had some conception of how the railway operated.
We would not have gotten any of the major safety-related absurdities of the modern railway.

 

[QueensCurve]

The line was already electrified and operated with both systems for a period of time, it was only electrified at 25kV for the benefit of freight operations.

I was used by The European passenger service for a short time in the mid 1980s.

Now that the 25kV estimates that claimed reelectrification woudl be cheaper than renewal have been shown to be "junk" - this idea is well and truly dead. Indeed even the Gibb report pointed out that electrification with third rail was likely cheaper than at 25kV and only regulatory problems stand in the way of the former.

Recent 25kV schemes have indeed been dogged by massive cost and programme overruns. It remains that, sensibly specified and competently installed, 25kV may become affordable again. The danger is that it will be so long before it happens that any experience regained will be lost again.

Also the opportunity to replace the 3rd rail will have been lost because the equipment will need replacement in any case.

This has to be weighed against the extra maintenance of all the extra pantographs and transformers on all the passenger units that would run over the converted section.

As noted above, most stock is being specified as dual current now so the "extra maintenance" is already committed.

Additionally a Class 92 can still develop over five thousand horsepower on third rail. [4000kW, coincidentally the same peak demand as a pair of Class 444s]

As against about 10,000hp on 25kv if my memory serves me rightly. I seem to recall they have a similar rating to the Swiss Class Re6/6.

Meanwhile AC systems would be fed with a magical supergrid bulk supply point that consumes no land at all

No, they won't, but they do require far fewer substations than with DC systems, particularly when "autotransformer" is used.

So in almost all cases the DC units will be able to move

In the case of the London Underground, stock can continue to operate if one of the two electric rails goes down. The voltage available, and therefore the speed of the train, is reduced.

 

[QueensCurve]

Perhaps someone needs some more 88s with third-rail shoes...

Oh no, please no.

 

[HSTEd]

I was used by The European passenger service for a short time in the mid 1980s.

Well yeah, but I very much doubt a tiny handful of passenger services could make a business case, even in the BR era.

As noted above, most stock is being specified as dual current now so the "extra maintenance" is already committed.

Blocks of concrete tend to require less maintenance than a pantograph, transformer, VCB and all that sort of kit.

As against about 10,000hp on 25kv if my memory serves me rightly. I seem to recall they have a similar rating to the Swiss Class Re6/6.

Well that's true - but freight operations on the southern region don't necessarily need that many horsepower, freight operations currently make do with far less horsepower than that.

By comparison a Class 92 on third rail still has as much power as a Class 90.

No, they won't, but they do require far fewer substations than with DC systems, particularly when "autotransformer" is used.

But the substations tend to be much larger in area terms.
Unfortunately I missed my PhD by 45 marks out of 3200 so I will not be doing the research I hoped on new technologies for DC railway electrification.... but maybe next year after my MSc.

The cabinets for a low power cooperative third rail substation are something like the size of an autotransformer unit and can be secured wherever convenient near the railway.

 

[gsnedders]

Recent 25kV schemes have indeed been dogged by massive cost and programme overruns. It remains that, sensibly specified and competently installed, 25kV may become affordable again. The danger is that it will be so long before it happens that any experience regained will be lost again.

I'm hopeful we won't see such a slowdown in Scotland, though it's not implausible depending on how much of the Scottish network ends up electrified the teams will have sufficient work to do in Scotland for a long while leaving nobody with experience free.

 

[Taunton]

There haven't been many lines that have been converted from DC to AC, but one was the North London Line, which had 3rd rail extended in the mid 1980s through to North Woolwich, and then much of the line converted to 25Kv about 10 years later. The actual costs and man-hours of the two sets of works must be still around, and it would be interesting to see them compared.

I recall actually seeing the works train laying the 3rd rail from Canning Town to North Woolwich, half double and half single track. I think it took them two weekends to lay it, the insulator pots having been previously fixed. No super "High Output" train, no super-Swiss project design offices, no big civils for bridge raising or whatever. Just one works train propelled by a Class 08. It wasn't even as if the Chesterton Junction, Cambridge, PW crew, who did the work, were that familiar with 3rd rail. In comparison to putting in 25Kv masts and wires, the time, effort and cost was a pittance.