Great Western Coffee Shop

http://www2.nationalgrid.com/UK/In-your-area/Projects/melksham-gwm-connections/


The Melksham electricity site is right alongside the TransWilts.

So they're already 50% complete?

The info I have ............. yes

National Grid are on program to deliver the 25kV connections as contracted ...............

You can also guess when NR start to pay the connection tariff whether they draw energy or not

So thats what all that work is for in Corsham Road, near the Whitehall Garden Centre, Lacock?

Grid outages especially ones on the 400kV National Grid are planned in 5 years in advance if a project misses the slot it can be a long wait.

I thought that the national grid supplied railway traction substations at high voltages, typically either 275KV or 400KV  and that a network rail substation transformed this down to 25KV or to a 25/50KV autotransformer supply for traction purposes.

This report suggests that the national grid are supplying the railway directly at 25KV via buried cables. This seems a bit unlikely, perhaps it is typo ?

The supply to the railway from the TNO (Transmission Network Operator) is at 25kV or 25/0/25kV (ie 50kV), the transformer and associated switchgear at the grid site is a "single user asset" and the railways pays for it when new and when it has to be renewed.  The maintenance is paid for through the tariff.

The electricity meter, and yes there is one its in MWh and MVArh, can be located either end of the feed cable depending on how the tariff was negotiated.

I assume this is a three phase feed. How does that work in the case of 25/0/25 kV?

Not answering Ellendune's question (over to you ET) but for general information...

25kV-0-25kV is a single phase centre grounded 50kV supply in the same manner as the 60V-0-60V UK building site transformers.  HS1 uses it and I have an article somewhere from around 10 years ago that describes it. On HS1, the opposite 25kV supply is carried parallel to the energised catenary outside of the pylon structures and the two are connected together at intervals by autotransformers.  The net effect is doubling the power available whilst reducing the voltage drop, and number of TNO supply points.

There are several reasons why modern 25kV grid feeds are taken from the 400kV system and not the 132kV system, one it does allow a higher power take,  the unbalanced loading of the railway single phase supply has less effect on the network (negative phase sequence) also the 400kV is strategically more secure.


In essence correct, it reduces the I²R losses however the fault level goes for a classic 25kV system it is 6kA but for the 25/0/25kV it is 12kA (although this can be reduced with use of reactors) which increases the I²T.   This means the conversion of a classic 25kV BT system to a 25/0/25kV AT system can mean modifications to the OLE registration arms.  The system designers also have to be conscious of the RoEP (Rise of Earth Potential) so bonding has a big focus currently in the industry.

Thanks ET I had assumed that a single phase feed would cause too much phase imbalance.  I assume that the Didcot feed will come from a different phase to the Melksham one to at least give the whole grid some better balance. 

Not bad I thought from a geologist!  ;D
In my industry there is a traditional - and constant - tension between geologists and engineers. I had to look up I²T, but then every day is an education.

To return to your phase supply, am I correct in assuming that with a three-phase grid system (say phases A, B & C) that the supply primaries at (say) Didcot would be A+B, Melksham B+C and 'Somewhere-else' C+A?

There are 3 Grid supply points on the English side of the Seven, Kensel Rise, Didcot and Melksham; all will have 2 Grid transformers the 2 transformers at each Grid site will be split across phase pairs eg one will be R - Y the other will be B - Y.

In the OLE there is an item called a Neutral Section, these have been up to now 2 GRP rods with ceramic beads the 2 rods separated by a short (about 500mm) of earthed contact wire http://www.google.co.uk/url?sa=i&rct=j&q=&esrc=s&source=images&cd=&cad=rja&uact=8&ved=0CAcQjRw&url=http%3A%2F%2Frailsimroutes.net%2Fblog%2F%3Fp%3D912&ei=O_dWVazBJOqw7AaOuYO4Bw&bvm=bv.93564037,d.ZGU&psig=AFQjCNFP5mc14ApAOXDua11_WMf5FrF9yA&ust=1431849145544698
There is a plan to use carrier wire Neutral Sections on GWEP this is a series of overlapping sections of contact wire where there are 2 floating sections and an earthed section.

The Neutral Section allows the train pan to pass from one electrical supply to the next, there is an Automatic Power Control system which opens the trains circuit breaker.

So the 50kV is between two phases and the 25kV/0/25kV is phase - neutral - phase (e.g. Red - Neutral - Yellow or Blue - Neutral - Yellow)?

On the 25kV system it is refereed to as +25kV and -25kV and the 0 is traction return.  The + and - from one transformer will be out of phase with its mate at the same Grid site and is likely to be out of phase with the one 50 miles away, even if it were to be in phase the railway would not parallel the 2 Grid sites' all sorts of weird power flows would happen.

Mid way between Grid sites the is a MPATS (Mid Point Auto Transformer Site) Maidenhead is the MPATS between Kensal Rise and Didcot so there will be Neutral Sections at Maidenhead MPATS.  It is also the boundary between 2 different protection schemes (by protection I refer to electrical over current etc) which need to interface with each other. Why different ones well the Crossrail scheme was developed and tendered a few years a head of GWEP, Crossrail will a similar AT protection scheme to that used on the WCML and GWEP is "new and novel"

So it is like this.

(http://upload.wikimedia.org/wikipedia/commons/3/3f/Onderdel_spoorwegnet.gif)

Source Wikipedia (http://en.wikipedia.org/wiki/25_kV_AC_railway_electrification)

The primary on the main feed transformer is between two phases and the return is connected to the centre tap in the secondary (which is also earthed?).

Sorry images defeat me follow the link ???

Admin Note: Link to image fixed - bobm

Polyphase terminology can be confusing. You might reason that three-phase has three voltages equally spaced over a whole cycle, so 120 degrees apart. So two-phase would be two voltages 180 degrees apart, right? No, that means they are 90 degrees apart. The transmission link from Melksham to Thingley does have two voltages 180 degrees apart, but that's still single phase.

The reason why is that each phase comes from a winding, in a transformer or generator. So each has two ends, and until you connect one end (or the middle if it's available) to something its phase is ambiguous - the ends are 180 degrees apart in phase. With three phases you can connect one end of each winding to a common point, almost always earthed to define the peak voltages and for protection purposes. Your transmission line can then convey all three phase voltages and, optionally, the common as well. But you can also connect the three windings in a ring, which gives three phases and no common point.

The point here is that every phase can also supply its antiphase, so the three phases are really 60 degrees apart by dividing 180 degrees by three. So two-phase is 90 degrees apart. In this case you can't connect them in a ring, you need four - because there is no two-sided polygon (think about it).

The idea of running phase and antiphase conductors plus a common is (or was) very common with DC. It's called a three-wire system (I think the AC one can be as well). Since you can't make a transformer for DC, doubling up is the most you can manage to raise the voltage for efficiency. London Underground used it a lot, and still may in some places.

It also happens in North America for domestic electricity supply, where the practice was set by Edison's DC system before Westinghouse and Tesla won that war. Most individual houses have a pole-mounted transformer, with a 120V/0/120V centre-tapped secondary. The two "hot" wires are red and black (from DC practice), and the centre is white and earthed at the pole (but not very well). Big fixed loads like heaters, aircon, cookers, are run off 240V between red and black. Socket circuits are run off one or other of the "hot" feeds and common, so should use red/white or black/white installed cable. And that's why US flex uses black/white - though why they chose that and not red/white I have no idea.

Installation practice seems to have been based on the idea that 120V is "safe" - so in our terms it is often scary. True it is safer, but a row of sockets above the bath, for example (I saw that in Canada) puts them in one place where 120V can certainly kill you. So more recently houses have to be protected by a green (or in cables bare) ground wire, usually going to a lump of buried metal. Sockets can also be protected by a GFCI (ground-fault circuit interrupter), though not (AFAIK) whole houses (yet).

The laws of physics and of physiology are universal, so the diversity of electrical wiring rules and practices is pretty amazing.

Certainly how my British kettle was wired during our five year sujourn in the States, plus all my workshop equipment that I took there with me.

Black is the normal live or 'hot' even on the anti-phase circuits - Red only appears in 3-wire+E cable. Black-white-green solves the colour blind problems associated with red. My challenge with each new house was to find the nearest source of (co-breakered for safety) 120V+120V=240V.  Sometimes it was a light fitting (!)*, sometimes I had to go all the way back to the DB.

*Americans do not differentiate between lighting and power circuits - that took some getting used to.  :-\

From my limited understanding of USA wiring, they retained the old dc colours ie + is red - black and 0 is white because they have 115-0-115 V system.

Back to real power stuff here in the UK.
One of the requirements now is to keep touch potentials below 50V and in the case of services (like station domestic supplies from the DNO) we are not allowed to export a rise in earth potential of 25V this is challenging for what is an "earthy" return system.   

So presumably there will be another MPATS around Swindon between the Melksham and Didcot Grid sites?  Also, what do the switching stations do that are going in at places like Pangbourne, Wantage Road and Uffington? Do these simply break the line into electrical sections or do they also provide a feed from the -25kV line?

The classic way the track has been sectioned in the UK on 25kV railways is a circuit breaker ever 7 miles for each track at both ends.  Circuit breakers in this form have been deemed to expensive so a system based on the French Railways of sequential isolators is being used, basically the track breakers every 7 miles have been replaced with isolators, the event of a fault the main circuit breaker will trip the protection scheme will know where the fault is (because each isolator has fault detection) it open the isolator and recloses the main breaker................. all this in literally seconds from fault detection to clearing faulty section to reenergise healthy sections.

The French do theirs manually at the control room in the UK we will be using RATS which is a BS EN 61850 (Communication networks and systems for power utility automation) based system

Well, sort of. This diagram from ABB (http://www04.abb.com/global/seitp/seitp202.nsf/0/8b5525f13b256037c1257c9e005562ed/$file/ABB+GWEP+Map.pdf), who are installing all this stuff, shows the plan. The English feeders are at Bramley, Reading, Didcot, Royal Wootton Basset, and Thingley Junction.

While ATS come in four flavours - plain (ATS), sectioning (SATS), feeder (ATFS), and mid-point (MPATS) - practice does seem to have evolved separately from terminology (again). So while an ATFS feeds one, two, or more sections from the end, and most sections have feeders at both ends, there exceptions to that. MPATS are now only for major boundaries, and within each section there are ATS and SATS. One SATS per section might be expected, but again that's not what happens. I suspect that sections are short enough that, in effect, the mid-point is replaced by the next feeder.

But Wales does seem to be different - three ATFS in a row, and a TSS (not sure what that does). Multiple feeders may be for security of supply, and happens elsewhere (e.g. the three on HS1 at St Pancras, Stratford, and Barking). But why are there MPATS at Filton and Westerleigh?

Obviously the actual design is based on many extra unknown (by me, anyway) factors. I'm still puzzling over how the AT feed arrangement allows you to replace the return conductor by a screen conductor, and how (or if) it acts to reduce earth return currents. I can see how a booster transformer does that - now there's another bit of fossil terminology; the last thing it does is "boost" anything!

AIUI the vast majority of the return current flows in the other (anti phase) 25 kV line.   

If for comparison you think of a 3 phase star connected supply with a balanced load, the neutral conductor is effectively carrying no current.

Very many thanks to stuving for the GW network circuit diagram and for ellendune's AT system diagram.

I'm surprised to see the extra feeder points as I understood these had to be taken from the 400kV supergrid at some expense or proximity to a substation with a spare transformer bay. Presumably there are some long buried cables. Where is the Reading feed to come from?

The weakness of the old system was that it couldn't cope with excessive traffic, with line volts dropping to 19kV at times, close to an undervoltage trip. Modern transformers (x4) allow for less regulation (i.e. voltage drop at the transformer). Ellendune's diagram shows how the imbalance of load being drawn across one side of the 25-0-25 system should be opposed by the AT, pushing up the OHL voltage and balancing the earthed neutral track potential. The problem there is that a lot of electrons have to flow from point 7 to points 10 and 4, raising the track voltage away from the earthing points at the transformers. I've read of this being measured at above 100V at times. Perhaps that's why there are quite a number of close  AT's.

Can't wait...

OTC

When the environmental studies were done, Wiltshire were told this:

I'm not sure if that's still the plan, or if they would have to bury a cable - it's about 22 km. Bramley and Didcot are both grid station sites, but as you say Reading is not. However Bramley is only about 14 km from Reading, depending on the exact feed station siting, so the same method should work here too.

This shows how much lower a Voltage drop/m a modern system achieves, even with higher current. The ECML feeders (at 25kV) could reach a maximum of 18 miles/29km, or 36/58 with a 45mph speed limit.

OTC

My understanding is the 400kV Grid supply to Didcot ATFS (Auto Transformer Feeder Station) was commissioned this weekend (8/9 Aug 2015) not heard if it was fully successful yet.  All that is needed now is the knitting strung up  ;D

Good to know National Grid aren't behind schedule!  :P

They may be: we just haven't heard about it yet.  :P ::) ;D

I noticed at Pangbourne a few days ago that some dangly bits had been fixed to some of the masts there. Only seen from the road under the Whitchurch Road bridge, so no idea how much has been achieved.

 'Dangly Bits'.

 Now that's the technical jargon I can understand !    ;)

Actually I think they are, quietly relived that NR don't want to take power for a while

These may seem like daft questions but with so much now appearing at the trackside I hope some of the experts here will be able to answer them.

1) There are four tall posts either side of the tracks just west of the old line into Didcot A Power Station. Are they part of the neutral section or is that nearer to Foxhall Junction?
2) Maybe these masts that will eventually supply power from Didcot ATFS to the OLE west of Didcot as far as Wootton Bassett Junction? There is a cable trough back to the Didcot ATFS, and another one to a similar set of posts near Didcot North Junction for the Oxford lines.
3) Does this mean Didcot ATFS provides a power source for three sections - 1) Maidenhead MPATS to Milton, 2) Milton to Wootton Bassett and 3) Didcot to Oxford?
4) Am I right in thinking there are only neutral sections where the source of power changes from one ASTF to another?
5) Does anyone have a drawing showing the general arrangement of a GWEP neutral section?
6) Finally, with an ATF system, does +25Kv phase flow through the OLE and -25Kv phase flow through the ATF cable and they come together at an ATS, get boosted, and then repeat that to the next ATS and so on?

5.   There was a pdf guide quoted in the other thread the other day,  http://www.railwaysarchive.co.uk/OLE/
It describes the GW series 1 OHLE as using a "carrier wire neutral section" on pages 88/90.  I think because there are overlapping electrically isolated sections of contact wire over a significant distance this explains why they look so complex compared to inline insulators.

6.  AIUI there is no boosting as such.  What the auto transformers do is to share the current 50/50 between the contact wire and the autotransformer feeder, (ATF).   Again, if you look at the section of the guide around pages 46/47 you can get an idea.   Because the current flowing in the contact wire and ATF are halved, then the transmission losses are significantly reduced because they are proportional to I².

Paul
 

1). Those sets of tall posts in pairs appear at all lineside sites so, even without ever seeing how one is wired, they obviously do make the electrical connection(s) to the OLE. At an ATS, the autotransformer(s) just connect to the catenaries and ATFs. At an MTAPS (e.g. Maidenhead) there is a section break, left open in normal operation, needing switchgear and AT connections both sides of it. I can see four post pairs on the London side and two on the Reading side.

An SATS is in theory the same as an MPATS, but with the break normally closed and only opened during fault/maintenance operation. An ATFS does not need a section break, or even ATs (e.g. Kensal Green). But for either of those at a junction, extra section breaks (and perhaps ATs) are needed. I have no idea where they are being put at Didcot or Reading, though there is more than enough ironmongery to make one around Kennet Bridge.

4). As implied above, you need breaks to separate power from separate feeds in all feeding modes that may arise. Only a few breaks are open normally, and how many others are provided will be the result of a "nice to have" versus "not on my budget" discussion.

5). When a pantograph transfers from one contact wire to the next, it connects them. You can't allow that at a section break, so there has to be an extra insulated wire run in between. How much length this takes is subject to a lot of variation.

6). There's no + and - involved, just two feeds of power opposite in phase (antiphase). That's like two pistons 180^o apart on the crankshaft, or the directions your arms point if you stick them out sideways and twirl round until you fall over. I think it's simplest to see the ATF as a second wire that can carry a second, equal, reserve supply of power to top up the line feeds. Between ATs, trains draw power and the line voltage drops. By the magic of magnetic induction, the autotransformer responds to that voltage drop by taking power from the ATF, inverting its phase, and topping up the line catenary wires.

This inversion of phase results in similar currents in the two overhead feeds, but in antiphase (they would cancel if added) so only a small current flows to earth via the rails. That's useful, as too much earth/rail return current is a nuisance. (I think I² and V² are red herrings here.)

The feeds to the OLE and ATF go on "across track" span wires from the OLE lineside switches to "drapes" that connect to the OLE  there are insulators segregate it all.  These are place well above the OLE to allow for a cross track span wire being alive when part of the OLE below it is isolated & earthed to work on

To allow for parts of the Didcot triangle to be isolated yet maintain a feed north to Oxford cables are run at ground level, in BR days "bare feeders" were used on the side of the OLE structures however this causes problems in "emergency" isolations because the bare feeders have to be switch off as well.
Yes that's the plan
It would be possible to feed Oxford from Wooton Basset in the event of a total loss at Didcot, however this would be second emergency feeding and would impose major timetable restrictions


There are Neutral sections ate MPATS eg Maidenhead and ATFS eg Didcot.  When Bramley is built there will be Neutral sections at Reading

The neutral sections are needed for two reasons, 1) often the supply to the Grid transformers are taken from different phase pairs to allow for some balancing of the loading across the Grid, 2) to prevent power flow between Grid sites via the Railway OLE system.

The GWEP neutral sections are what are called carrier wire neutral sections, basically 3 over lap section of OLE, a floating section , an earthed section, a floating section.  The pan transitions form the live wire to
the floating section, then the earthed section to the floating section and onto the live section.

There were trials done to see if the earthed section could be done away with, not sure of the results of this.  The carrier wire neutral section does have by pass switches in the event a train gets stranded in a floating or earthed section


The + 25kV and - 25kV terms were used in the early days of the WCML conversion to AT to make it easy to explain to the lay railway(wo)man.   The system is 25 - 0 - 25kV the train only ever see 25kV but the power is distributed at 50kV there by reducing the I²R losses

Thanks chaps my head hurts less now.

Mine hurts more  ::) ::)

Thanks to Electric train, stuving and paul7755 - I knew you guys would come up trumps!  Thanks for explaining it all - its slowly sinking in.

Somebody had to invent all this stuff from nothing but a few equations, maybe developing a few more equations on the way!