or maybe put juice rail in to the first or second station after hurst grn. Less time on battery, but no new feeders needed.
-
Our booking engine at tickets.railforums.co.uk (powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!
Alternative To Third Rail?
- Thread starter Doomotron
- Start date
- Status
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R
RailUK Forums
Yes, easily. Although a short stretch of con rail Uckfield - Buxted (with no back up) would cover just about every contingency.
Marshlink and North Doens would be even more simple.
a static area like a station or siding is much better as you need a substation every 3-5 miles
also advantage off a static charging via local 3rd rail in a siding or platform is the ability to turn it off easily locally or remotely to reduce the electrocution risk in a remote area
also advantage off a static charging via local 3rd rail in a siding or platform is the ability to turn it off easily locally or remotely to reduce the electrocution risk in a remote area
Last edited:
DerekC
Established Member
Rather than start a new thread I thought I would revive this one. Rather than an alternative to third rail, how do we get ORR to climb out of the trench it has dug itself into on local extensions?
The suggestion of battery/third rail vehicles in this thread is one way to avoid fitting high risk locations (like level crossing approaches and perhaps stations and yards). An alternative would be to have switchable sections for high risk areas which are only energised as a train approaches. It wouldn't involve lots of expensive switchgear because you don't have to switch current. A motorised hook switch would do the job, driven from the signalling system so that the switch closes when a train enters the area and opens again when it has left. I suspect that would be difficult to modify existing this rail installations, but it shouldn't be too hard if electrification is being extended. Maybe even link it to level crossing strike-in points.
The suggestion of battery/third rail vehicles in this thread is one way to avoid fitting high risk locations (like level crossing approaches and perhaps stations and yards). An alternative would be to have switchable sections for high risk areas which are only energised as a train approaches. It wouldn't involve lots of expensive switchgear because you don't have to switch current. A motorised hook switch would do the job, driven from the signalling system so that the switch closes when a train enters the area and opens again when it has left. I suspect that would be difficult to modify existing this rail installations, but it shouldn't be too hard if electrification is being extended. Maybe even link it to level crossing strike-in points.
Last edited:
Chiltern006
Member
- Joined
- 3 Oct 2018
- Messages
- 657
With the third rail fiasco, I think it looks a lot better than OHL. It looks part of the line. Also I think they should electrify the Chiltern Mainline with it as bridges are too low as well as the Exeter Line on SWR, and the SN diesel lines. Possibly even the North Downs Line.
HSTEd
Veteran Member
- Joined
- 14 Jul 2011
- Messages
- 16,722
Well other than telling the ORR to actually read the Electrical Regulations it repeatedly cites to support it's position?
The regulations specifically declare railway installations to be out of scope, along the lines of things like aluminium smelting plants and other major industrial facilities specialising in electricity use
It is just more of the gold plating obsession that ORR pursues because it wants a "proper railway" like the other kids on the continent have.
That also led to the idiocy of trying to insist on multiple pantograph 125/140mph operation even though that has never been necessary in the UK and was only necessary to save a handful of additional IEP vehicles.
And the abandonment of specailised UK voltage clearances.
You fit the centre of platform lines with normally dead sections that are only energised if the train has lost a shoe and thus is gapped.
The sections would be set to be just short enough that a typical train will have one shoe on the live rail at the ends of the platforms, but if not the driver/guard can press a plunger that will liven up the rail to enable the train to take current normally.
Solid state switches are really cheap now.
Almost all non-trespass deaths and injuries occur at platform interfaces anyway. And tresspass deaths can probably be reduced significantly by just turning the supply off when the line is not in service - although with modern cooperative rectification that becomes more difficult.
Last edited:
DerekC
Established Member
All good stuff - but what's co-operative rectification? I Googled it and the only thing that came out was something to do with Brownian motion in confined ratchets - so I was none the wiser!
hooverboy
On Moderation
- Joined
- 12 Oct 2017
- Messages
- 1,372
rectifier technology is becoming interesting now!
lots of design/development work going on in power gallium nitride semiconductors.
the prototypes they built look to be somewhere in the order of 40-50% less power dissipation than current rectifiers,
at several KW each on typical rectifiers being thrown off in pure heat,this looks like a solid improvement.means smaller components, less weight, easier to fit, less cooling required.
HSTEd
Veteran Member
- Joined
- 14 Jul 2011
- Messages
- 16,722
The traditional rectification model is to have a relatively small number of multi megawatt substations.
See this on the southern regional and elsewhere.
But there is a streetcar line in the US that uses 300kW or so rectification stations to support the line, positioned much more closely together. THe advantage is the systems can be fed from low voltage supplies which drastically reduces the cost of the substation and makes them much more compact as all the HV safety gear can be dispensed with.
A modern rectifier is also not a dumb device like a silicon or mercury vapour diode - it can control how much current it allows to flow and so on.
So you could imagine a chopper-rectifier that produces its rated output power into the conductor rail whatever the voltage on its output sits at, but allows the voltage to sag if too much power is being drawn rather than simply tripping out.
This causes power to move along the conductor rail/line from adjacent substations, which then cooperate to ensure that demands are met, and since resistance losses scale by the square of current, providing only a little power close to the train is extremely advantageous. IMagine spacings of 400m or potentially even less!
Modern hall effect sensors would allow each substation to measure the current flowing along the rail, allowing them to still trip out on an overload, they would then signal the substations down the line to trip out as well - or they would in turn be overloaded and trip if the intertripping failed.
Since the substations are much smaller there is no need for DC-side circuit breakers as the controlled rectifiers can simply turn off if necessary, or open their AC side circuit breakers (Which are much cheaper commodity items, especially in LV) in an emergency.
With more complex substations (which can be derived from developing solid state transformer technology) we could also make the substations hold the rail voltage at about 850V, and start withdrawing power if the rail voltage floated higher than that, which will allow regeneration into the power grid.
Meanwhile the trains are pretty much none the wiser and thus it is backwards compatible with the existing stock.
apk55
Member
Substations are not cheap. For a start you require a plot of land to which you can bring in high voltage cables and crane in large heavy modules. Hence most are located at or near stations where there is easy road access and land available. This may be a problem in many urban areas as the railway is closely bounded by residential property for long lengths and rural areas surrounded by fields with no easy access. Therefore substation spacing is often less than ideal.
I very much doubt that semiconductor switches would be considered safe for isolation purposes as they can leak current , typically up to a lethal 100mA is possible. If there a safety isolation requirement then normally a double break contractor should be used as contacts can be known to weld under fault conditions.
I very much doubt that semiconductor switches would be considered safe for isolation purposes as they can leak current , typically up to a lethal 100mA is possible. If there a safety isolation requirement then normally a double break contractor should be used as contacts can be known to weld under fault conditions.
HSTEd
Veteran Member
- Joined
- 14 Jul 2011
- Messages
- 16,722
The whole point of using more small substations is there would be no heavy modules and no HV supply cables!
Indeed SSTs in this power range are under development that could be mounted on a utility pole in the American style!
But these are not conventional substations, they aren't that much bigger than an ATF transformer would be, and potentially even smaller!
5000A current isolator, which is massive overkill for platform disconnection, would require something like 80 parallel connected SiC MOSFET modules like this one, and that is overkill since it would be continuous thermal rating.
Typical drain current leakage is something like 2uA per device outside fault conditions, which would be something like 160uA leakage, which is negligible given the none-zero leakage through the rail insulators etc etc.
Whilst they can leak 100uA per device, that would still only produce 8000uA or about 8mA, even if all the devices were at the very top of spec, which we can easily monitor and replace modules that climb out of an acceptable range.
Also this is not isolation in the safety critical sense, this is risk reduction in that the rail would still be considered live at all times whilst this system was in use.
What fault conditions could reasonably cause contact welding in this system?
The system will be interlocked to prevent it opening whilst significant current is flowing through it?
And again, this is not a safety critical system because the third rail is still to be considered live at all times unless the master supply is set to be isolated.
Also such a fault would be easily detectable by the interlocking system and could be reported to the power control room for rapid replacement.
Elecman
Established Member
The cost of 300kw DNO supplies in most areas would be somewhere between £30k to way north of £200k per supply site, depending how much reinforcement of the local network is required ( or the DNO planners think they can get away with charging your schemes for reinforcement that they should already be doing for voltage regulation or resilience reasons) so if your proposing these substations even every few km it’s a non starter financially.
apk55
Member
As a minimum you would be looking at a substation rating of 1MW which is only the power of a double unit Manchester tram. The most powerful pole mounted transformers are around 50 to 100KW an order of magnitude less. The substation for my estate (100 houses approx) is the size of a large container.
You would still need power feed to the substation from the local 11KV distribution network which nay be some distance away and could involve digging up roads inquiring wayleaves etc.
Semiconductors have a habit of failing short circuit
You would still need power feed to the substation from the local 11KV distribution network which nay be some distance away and could involve digging up roads inquiring wayleaves etc.
Semiconductors have a habit of failing short circuit
HSTEd
Veteran Member
- Joined
- 14 Jul 2011
- Messages
- 16,722
Previous calculations I have done suggest that for many routes, a substation spacing of about 1000m would be sufficient for a twin track railway.
£200k per twin track-km.....
25kV costs £3m/route km and up on such a route.
So it's not really a non-starter at all.
Again, did you actually read anything I wrote?
A substation does not have to be 1MW because they can share the load using modern power electronics.
But this is not a line frequency transformer - it would contain a medium frequency transformer because it is more akin to an overgrown switch mode power supply.
Where the DNO put it's substations is no concern of the railway.
The DNO would handle the supply of about 300kVA of 400V supply to the substation, and how they deal with it is there concern.
And?
What are the consequences of this?
The fault would be detected immediately and repair scheduled.
Even if it only functions 90% of the time, the number of FWI caused by people falling onto the rail in the platform would be reduced by 90%.
EDIT:
I would also suggest applying a polymer/silicone rubber insulating coating onto the side and bottom of the rail, with small patches removed for cable attachments etc.
That way if someone is trespassing and drags their shoe against the rail they are far less likely to get a substantial shock.
Last edited:
apk55
Member
There is no way the 415V local supply network could supply a traction load.
For a start you would need to run a heavy duty cable back to the substation as the cables in an average street would overheat with anything more than 10s of KWs extra load. Then the substation which often have a continuous rating of a few hundred KWs would need upgrading to cope with vast increase in load. Finally as domestic supply substations are typically spaced about 500M apart and you could only put one traction supply per substation area then you would be lucky if you could supply power at even 1MW per KM of track - barely enough to run a very light tramway service.
For a start you would need to run a heavy duty cable back to the substation as the cables in an average street would overheat with anything more than 10s of KWs extra load. Then the substation which often have a continuous rating of a few hundred KWs would need upgrading to cope with vast increase in load. Finally as domestic supply substations are typically spaced about 500M apart and you could only put one traction supply per substation area then you would be lucky if you could supply power at even 1MW per KM of track - barely enough to run a very light tramway service.
Railwaysceptic
Established Member
- Joined
- 6 Nov 2017
- Messages
- 1,409
I haven't read the Act Of Parliament the ORR rely on to justify their position. Does the clause concerning railway installations refer only to existing track and equipment or does it also include anything new?
Agreed - completely, plenty of dreaming going on.
The current minimum power assumption per train is 2.7MW, hence the capacity of substation is dependant on the density of trains in the supply area (with larger supply areas being better for regeneration but worse for everything else). The maximum current flow at any point in the circuit is limited to 9kA so nominally 6.75MW which is enough to support 2x 12 car 700s on max 3rd rail performance setting.
In reality the cost estimate from the DNO for doing that would make it cheaper to install your own 33kV cable which is why L&SWR + SER / SR / BR / RT / NR have always done that!
The 3rd rail substation installation and upgrades is already virtually all done by contracting teams from UKPN and SSE (the 2 DNOs in the 3rd rail area but they will work on each others patches and everything is competitively tendered).
The biggest 3rd rail supplies are virtually effectively fed direct from Nat Grid at New Cross (ex Gas Works) / Plough Lane with rail exclusive SGTs and minimal DNO involvement.
All this dancing on the head of a pin to find a way to build more 3rd rail. For anything other than genuine metro lines with very controlled track access, the new 3rd rail ship has sailed. In terms of safety, energy use and performance, 3rd rail is a 100 year old solution that doesn't fit the requirements of a modern railway. Just like the solution to the NI border there may be some whizzo tech solutions in the pipeline that address one or more of the ORR issues, but in the meantime, OLE is the proven answer for anything being planned now. Despite a late start, nearly 2/3rds of all electrified track miles is with OLE. The continuation of 3rd rail use is under grandfather rights which includes replacement and minor additions that enable it to continue to function.
New and replacement bridges are generally built to current standards so when those over 3rd rail routes are time expired they will gradually become more compatible with the requirements for OLE.
edwin_m
Veteran Member
Except perhaps on lightly-used lines - there will always be some need for on-board power for those rural routes where the benefits of any form of electrification don't justify the capital and maintenance costs. And in rural areas power supplies sufficient to power a train are likely to be difficult to find regardless of voltage - but 25kV probably still has the advantage in only needing one every few tens of miles. I note also that the ECML upgrade is using supplies with inverters, which get round the phase balance issue.
On the Southern network there are only a few obvious infills (Ashford-Hastings, East Grinstead and maybe a couple of others). These probably aren't big enough to be 25kV islands but it's reasonable to expect the massive investment in battery technology to deliver at least the fairly modest improvement that would be needed to allow battery operation of these sections. Beyond that we're into major extensions of the south-of-Thames network such as the various routes to Salisbury, but I suggest this would be a 25kV scheme because of the distances, thin-nness of local power supply and likelihood that further extensions would interface with 25kV on other routes. There's a bit more scope for extensions on Merseyrail to Warrington, Wigan, Skelmersdale or Preston but these would all encounter 25kV routes (maybe not Skelmersdale) so 25kV is a realistic alternative to battery here.
So actually I think the scope to extend the third rail network is quite small, unless we are talking about radical proposals to electrify new areas with third rail where OLE would be the normal choice. And in that case we probably aren't interested in compatibility with the 750V top-contact system so could use something like 1500V bottom contact, which is ORR-friendly as well as being a bit more efficient.
Railwaysceptic
Established Member
- Joined
- 6 Nov 2017
- Messages
- 1,409
First, you haven't answered my question which was a genuine enquiry and not an attempt to find an excuse to install more third rail anywhere. Second, there are short sections of unelectrified track in ex Southern Region territory where third rail, far from being a ship which has sailed, could provide a fairly low cost improvement.
The ship which has sailed is the idea that the entire third rail system should be replaced with OLE.
Last edited:
HSTEd
Veteran Member
- Joined
- 14 Jul 2011
- Messages
- 16,722
Then why do DNOs quote for 300kVA 415V supplies as a matter of course then?
Which again, is done as a matter of course by DNOs - who operate enormous LV distribution systems already.
So the unit substation would be replaced by a slightly larger unit substation?
1MW per km of track is a ludicrously heavy traction load for most areas?
A twin track railway with 1MW per route kilometre would have Class 319s at 2km spacings in both directions, or Class 444s at 4km spacings in both directions.
The lines that currently remain unelectrified would have traffic densities far lower than that!
So which currently unelectrified railways would require performance of multiple 12 car 700s at maximum power output?
Also that implies single ended feeding! Otherwise the current rating is potentially up to 18kA, but slightly less as the circuits won't perfectly share the load.
A far more reasonable load is a single unit Class 319 or a Class 444.
Maybe a double Class 444 on the West of England main line but they are only hourly (half hourly Salibsury-Basingstoke)!
Southern Railway/Southern Region did not always use a 33kV distribution system they owned, in fact they later abandoned that concept in favour of grid substations connected to whatever supply was most convenient - which is how we ended up with 11kV connected substations.
This is just tkaing that concept to it's logical conclusion.
The SOuthern Region distribution system exists as a historical hangover from before the era of widespread high power grid systems existing.
So how is this relevant to the idea of electrifying new areas?
No areas that are not currently electrified are going to see any traffic densities like that.
And 25kV is a 60 year old solution that doesn't fit he requirements of a modern railway?
It is hilariously expensive
The only thing OLE is a proven solution for is burning vast sums of public money.
The only thing insisting on 25kV for new installations will achieve is ensuring there are no more installations
Except all the work of 30 years was blown up because the ORR decided it wanted a proper railway that had proper clearances in yet another example of mad gold plating.
EDIT:
For example, it takes roughly 40 minutes to travel between Basingstoke and Salisbury at the present time, assuming we take no credit for replacing 159s with 444s (as a traction exemplar for what a long distance third rail unit might look like), we would have at most two trains in each direction at any time.
Four trains total in the scheme, composed of 2 units each so 8 444s.
That is a maximum traction load, assuming all units take full power at once, of something like 16MW.
16MWe on the roughly ~52km between the SWML junction and Salisbury.
So that is a maximum possible traction load of ~300kW/route-km.
And that takes no credit for support from substations further west, or timetabling, or electronic power restrictions if the voltage on the rail sags (which units are required to do according to the specification), or any diversity at all.
Class 444s have such better performance than Express Sprinters that its entirely possible the number in trains in transit will never reach 4.
Last edited:
apk55
Member
For a new load of 300KVA you would almost certainly be looking at a new substation fed from the the 11KV grid. Many factories and industrial units with power demands like this usually have there own substation and a new large (50+) housing development probably would as well. And you would need somewhere to put it where you can easily route HV cables to. Might as well go direct to 750V DC from 11KV.
The 1 MW per KM is the upper limit of my estimation. In areas of lower housing density this may not be possible and in other areas where the existing network is stretched to the limit you may need laying in new HV feeder cables etc.
Limiting the train power to 1MW or less and preventing trains from running close to each other would lead to an inflexible system that would be barely worth doing
The 1 MW per KM is the upper limit of my estimation. In areas of lower housing density this may not be possible and in other areas where the existing network is stretched to the limit you may need laying in new HV feeder cables etc.
Limiting the train power to 1MW or less and preventing trains from running close to each other would lead to an inflexible system that would be barely worth doing
Last edited:
because that is what supplying local domestic /light commercial needs, that does not mean it is the best or cheapest way to supply a railway.
Which is why the capacity of 3rd rail substations has increased over time and track paralleling hut have been removed and replaced with substations etc.
Real timetabling assumptions won't be helpful and allow optimism like that
Uckfield if the 171 get replaced...
Indeed it also included 66kV! The cabling from the 11kV substations was still mostly BR/RT/NR's not the DNOs as cheaper for the railway. The 11kV supplies are busily being removed and replaced with 33kV to allow train lengthening.
In the Uckfuield case the local 11kV and 33kV supplies are awful with not ability to cover rail traction requirements so bigger longer distance feed is required. The 1980s ultra cheap south of Tonbridge 11kV solution isn't even available. The supply solution for Uckfield is Dormanland NG supply point just North of East Grinstead and run 33kV cable
The available supplies in large parts of the country are still quite bad especially where there are National Parks and AONBs etc surrounding areas including South of the M25 apart from along the South Coast and the BML corridor.
The Cardiff Valley supplies are very good apart from one valley which is getting tram'd
Freight when diesel is restricted in 20+ years time
BR tried that in the 1980s and NR have gone back to 33kV...
HSTEd
Veteran Member
- Joined
- 14 Jul 2011
- Messages
- 16,722
So we shouldn't design infrastructure to fit the outputs we want?
We should just hugely overengineer things without considering the actual needs of the route in question?
You should start with the desired timetable and all infrastructure improvements should be based upon it.
The unelectrified railway journey length there is 43 minutes, and two hourly trains in the peak so only a handful of formations will be in transit. An absolute maximum of four. And obviously since much of the line is single track they must be definition be quite spread out.
The fact that something is being replaced does not mean it was ever a mistake to install it at the time.
We might have to replace LV supplies with 11kV or higher ones in the future, but using the best solution for the traffic actually projected now is the best solution.
Improvements to the supplies in many areas are likely to be more affordable than building an entirely separate railway distribution system. Especially since many reasonable electrification projects are not going to use the capacity of even a single 33kV feeder.
Ah yes, the magical "freight will go electric, really it will" argument.
Freight have made no progress in switching to electric traction since privatisation, why would it change now? It has too much political backing for anyone to try and force them to do anything.
Also we should design infrastructure now for demands that might exist in 20+ years time?
Substation uprates, especially with this topology are not particularly disruptive affairs, as the new substation can be built next to the old or at any other convenient site and then simply be spliced into the main bus.
If the intertripping signals are properly connected there is no need for any of the existing infrastructure to be modified to enable this to work. Indeed the old substation would likely remain in service for the remainder of it's service life.
So we can add new substations as and when they are required.
superkev
Established Member
I quite like docklands bottom contact system with insulated top and sides which is used in a few metro lines in the world.
I wonder what the position of this is with the various regulators.
319s on the Windemere branch would seem an ideal application once they find out how much Hydrogen really costs.
K
I wonder what the position of this is with the various regulators.
319s on the Windemere branch would seem an ideal application once they find out how much Hydrogen really costs.
K
No I'm suggesting engineering for the requirements of the route not massively under-engineering which you are assuming to help self justify your thinking.
The entire SWML timetable is still limited by decisions taken planning during the Bournemouth electrification project 54-55 years ago so best not to repeat with similar under engineering in the future?
In many locations the only need for DNO network enhancements would be rail electrification so the cheapest way to pay for it is doing it directly especially when you see the DUOS charges as well.
Given that everyone else with experience doesn't agree with your analysis might suggest some more detailed recalculation is needed?
Windermere is so short that battery is ideal and the first mile or so of the branch (uphill) has no overhead obstructions and isn't in the National Park so a short OHLE extension would be easy and reduce battery duty cycles.
- Status