Overhead Line Power Capacity

[ryan125hst]

New rolling stock has been rolled out in many areas in recent years. New trains often have distributed traction with greater horsepower than the trains they replaced. When you combine this with more electric trains being introduced and, on some routes, the introduction of air conditioning when trains that previous operated the route only had opening windows, and the electrical load will clearly have increased.

Taking my local station of Retford as an example, instead of a mixture of diesel powered HST's and electric Intercity 225's, we now have electric and bi-mode Azumas meaning, if all is well, all of LNER's trains will pass through using electric power. Hull Trains now also use bi-mode trains, and Lumo have commenced service too, also using electric traction. It has made me wonder, as an Electrical and Electronic Engineer, what the capacity of each OLE section is on the ECML and indeed elsewhere in the UK. While it will be engineered for typical timetables (presumably as designed in the 1980's, although I believe power supply upgrades have been carried out across much of the ECML in recent years), what happens during disruption when multiple trains stop and then start to accelerate shortly after one another?

Looking at the Sectional Appendix, the overhead line section through Retford starts just north of the station, and the next Overhead Neutral Section (OHNS) is at South Muskham, about 3 miles north of Newark North Gate, around 16 miles of double track in one section in total.

So let's say there is disruption and there is a 10 car Azuma on platform 1 and another on platform 2 (using 10 cars in my example due to three motor vehicles per set, six per 10 car train). Then another Azuma stops on the up fast line to allow the late running service on platform 1 out, and a Hull Trains unit stops in the section behind it awaiting platform 1. If the train on platform 1 sets off and the section becomes clear, the Azuma on the fast line can now set off, meanwhile the Hull Train can now enter the now available platform 1. If the Azuma on platform 2 sets off just as these three trains are setting off, then before it reaches just north of Retford and thus the next overhead section, we will have four trains accelerating at once. In addition to this, there could be other trains heading north also drawing power. This is just a theoretical example but having seen trains bunch up one behind the other, and be sent to the platforms at Retford during disruption, this scenario is possible.

With all this in mind, what is each overhead line section capable of supplying, in Megawatts? Is it fairly standard for each section, or does it vary across the country? Looking at the Sectional Appendix it seems they are a lot closer together south of Peterborough, which suggests the limits of what can be supplied can be reached and therefore more sections are required to satisfy the power demand.

What is the limiting factor of the OLE power? Is it the substation and therefore substation upgrades allow greater capacity to the section. Or can the maximum current of the OLE by reached quite easily in busy areas, meaning the only way to increase capacity is to split it into a greater number of sections?

 

[GraphitProjekt]

I don't have the answer to most of these but I can say I don't believe there is a generic rating for each section. The early planning process of an electrification project starts around a prospective timetable for which the whole system is roughly based around. But there is meant to be some spare capacity to deal with unfortunately timed accelerations like that. I'd be interested to know if trying to stagger train movements for the reason of avoiding a trip is ever actualised in disruption recovery. I'd imagine not, quite enough things to be thinking about. There are also seemingly weak spots that tend to fail almost predictably. The very south of the ECML between Alexandra Palace and Kings Cross tends to get overloaded (notably recently) when lots of trains start moving after there's been disruption.

That being said, I believe power supplies can be expected to provide over their rated capacity for at least a short time, and (as previously said) there is supposed to be headroom for inconveniently timed loads and to cover for a supply loss. The 'average' feeder station has two grid supplies for each direction away from the site's neutral section. For a 132kV connection it seems to be in the region of 18 miles from my observations. But each of those supplies can be extended to cover the other one's load if that other one goes out (N-1 feeding)and support the timetable. So you could say that each of those supplies is only doing 50% ish of what it could theoretically handle - but actually it's at max capacity because of the requirement to cover for its neighbour.

Because of this, the ECML actually was able to swap HSTs for electric trains quite a long time before the ECML Power Supply Upgrade (PSU) was actually finished. This is at least a little because you can run things a bit closer to the edge and hope a bad set of coincidences don't happen. But I should say, I think this was mainly achieved on ECML by converting from RC/BT feeding to RSC feeding which could be done relatively quickly/easily and removes a non-trivial amount of built in impedance. In any case, ECML PSU was implemented and (mostly) finished (apart from the bit they abandoned...)

South of Peterborough there are definitely more trains so one would definitely expect a higher density of feeder stations. However, counting neutral sections is not always a reliable way of ascertaining the feeding arrangement as there are often "alternative" midpoints that are normally bridged over but only come into play in certain N-1 situations and there will also be feeder stations without neutral sections if they only have one feed present: there are examples of all of those on ECML south as well as random stretch of 2x25kV AT feeding between Hitchin and Welwyn.

 

[Dr Hoo]

The potential future loading from electric freight is far greater bearing in mind that we expect conversion from diesel on main lines and a 75% increase in volume by 2050. The new Class 99s can operate in multiple with a combined 'power' of 16,540hp apparently. I appreciate that power figures for locomotives are something of a black art but simplistically that is equivalent to 12,334kW. Imagine a couple of double headers re-starting from rest with 750m loaded intermodal trains either side of Stoke Tunnel one night after a stop for line examination or something. Could be getting on for 50MW for a short period.

 

[GRALISTAIR]

Either way 400kV supergrid connection to AT feeding gives the best performance.

 

[GraphitProjekt]

Oh yeah, perhaps useful to mention. The generic rated capacity for the system at large is given as a fault current of 6kA for standard (aka Classic feeding) or 12kA for 2x 25kV AT feeding with Supergrid Connections. I guess that's roughly 150kW and 300kW respectively

 

[HSTEd]

Oh yeah, perhaps useful to mention. The generic rated capacity for the system at large is given as a fault current of 6kA for standard (aka Classic feeding) or 12kA for 2x 25kV AT feeding with Supergrid Connections. I guess that's roughly 150kW and 300kW respectively

A fault current of 6kA is a fault power (assuming a resistive fault and a perfectly rigid power supply) of 150MW, not kW.

 

[The Planner]

Electrification projects not putting in enough grid feeders is a large part of the problem. They are very expensive, so if you can leech off an existing one that tends to happen. It kicks the can down the road for another project basically.

 

[172007]

Electrification projects not putting in enough grid feeders is a large part of the problem. They are very expensive, so if you can leech off an existing one that tends to happen. It kicks the can down the road for another project basically.

Lickey Bank being a classic example. Isn't the feed from New Street.

Affectively prohibits except overnight class 93's and 99's from using electric thereby not requiring a banker.

 

[hwl]

The new Class 99s can operate in multiple with a combined 'power' of 16,540hp apparently. I appreciate that power figures for locomotives are something of a black art but simplistically that is equivalent to 12,334kW.

99s are effectively adhesion limited up to 30mph which limits current draw.

The default maximum current draw per train on NR is 300A which is nominally 7.5MW at 25kV and less if the line voltage drops.
A 99 can draw ~7.2MW at the pantograph and all the Stadler multl-mode locos (88, 93 and 99) have:
i) current limiting to 300A for multi-working hence 2x99 will draw not much more more current than just 1 Cl 99 at medium and higher speeds with the current being equally split between the two locomotives.
ii) tractive effort limiting to ensure the applied TE is not greater that the coupler rating between the second loco and the 1st wagon there is no coupler snap, this would also keep the power draw at low speeds low for two locomotives.

Hence a pair of 99s is only going to take nominally 7.5MW not the 12.34MW you mention which is based on power at the wheel not total power (there has been some rounding and now using enough significant figure in the conversion factor when someone has converted the original kW figures to hp figure you quote hence it doesn't come back to the original kW number up on reconversion)

 

[Snow1964]

Class 99s are restricted to fit Network Rails nominal 7.5MW limit

Stadler makes a EURO9000 version (name comes from 9000kW), with 500kN of tractive effort

EURO9000 locomotive | Stadler

Leading the way to a more sustainable future, the Co’Co’ locomotive can be operated with up to three different drive systems: electric, diesel and/or batteries.

www.stadlerrail.com

From memory HS1 and the under construction HS2 have higher limits, some of the latest TGV (dasye) are 9280kW, so when operating in pairs could draw nearer 18.5MW

 

[hwl]

Class 99s are restricted to fit Network Rails nominal 7.5MW limit

Stadler makes a EURO9000 version (name comes from 9000kW), with 500kN of tractive effort

EURO9000 locomotive | Stadler

Leading the way to a more sustainable future, the Co’Co’ locomotive can be operated with up to three different drive systems: electric, diesel and/or batteries.

www.stadlerrail.com

From memory HS1 and the under construction HS2 have higher limits, some of the latest TGV (dasye) are 9280kW, so when operating in pairs could draw nearer 18.5MW

Yes higher limits on HS1/Eurotunnel.

On high speed lines the nominal maximum pantograph power limit is 10MW above low speeds (2MW at stationary / low speeds to avoid welding the graphite composite strip to the wire).

The theoretical peak power draw of a Cl.374 is just over 18MW but this is via 2 pans.

 

[thecrofter]

Lickey Bank being a classic example. Isn't the feed from New Street.

The single feeder at Winson Green (known as F4) feeds Cross City South and to the Neutral Sections at Aston and Stechford. In the event of the loss of Winson Green (N-1), Galton Junction (F1) is the alternative feed. Birmingham New Street is only a TSC.

 

[GraphitProjekt]

A fault current of 6kA is a fault power (assuming a resistive fault and a perfectly rigid power supply) of 150MW, not kW.

I don't even know how I cocked that one up! Guess I'll have to retake primary school maths at some point lol

Some more numbers I've come across
- average rating of a 132/25kV supply is between 25-26.5 MVA, but can be more like 18MVA on the low end
- supergrid connections whether to 25kV or 50kV is between 80 and 104MVA

 

[Class 170101]

Electrification projects not putting in enough grid feeders is a large part of the problem. They are very expensive, so if you can leech off an existing one that tends to happen. It kicks the can down the road for another project basically.

Effectively the GOB is limited to two electric freights at any one time as power is fed from off route feeder stations Acton Lane (Harlesden) and West Ham.

 

[Richard123]

New rolling stock has been rolled out in many areas in recent years. New trains often have distributed traction with greater horsepower than the trains they replaced. When you combine this with more electric trains being introduced and, on some routes, the introduction of air conditioning when trains that previous operated the route only had opening windows, and the electrical load will clearly have increased.

Taking my local station of Retford as an example, instead of a mixture of diesel powered HST's and electric Intercity 225's, we now have electric and bi-mode Azumas meaning, if all is well, all of LNER's trains will pass through using electric power. Hull Trains now also use bi-mode trains, and Lumo have commenced service too, also using electric traction. It has made me wonder, as an Electrical and Electronic Engineer, what the capacity of each OLE section is on the ECML and indeed elsewhere in the UK. While it will be engineered for typical timetables (presumably as designed in the 1980's, although I believe power supply upgrades have been carried out across much of the ECML in recent years), what happens during disruption when multiple trains stop and then start to accelerate shortly after one another?

Looking at the Sectional Appendix, the overhead line section through Retford starts just north of the station, and the next Overhead Neutral Section (OHNS) is at South Muskham, about 3 miles north of Newark North Gate, around 16 miles of double track in one section in total.

So let's say there is disruption and there is a 10 car Azuma on platform 1 and another on platform 2 (using 10 cars in my example due to three motor vehicles per set, six per 10 car train). Then another Azuma stops on the up fast line to allow the late running service on platform 1 out, and a Hull Trains unit stops in the section behind it awaiting platform 1. If the train on platform 1 sets off and the section becomes clear, the Azuma on the fast line can now set off, meanwhile the Hull Train can now enter the now available platform 1. If the Azuma on platform 2 sets off just as these three trains are setting off, then before it reaches just north of Retford and thus the next overhead section, we will have four trains accelerating at once. In addition to this, there could be other trains heading north also drawing power. This is just a theoretical example but having seen trains bunch up one behind the other, and be sent to the platforms at Retford during disruption, this scenario is possible.

With all this in mind, what is each overhead line section capable of supplying, in Megawatts? Is it fairly standard for each section, or does it vary across the country? Looking at the Sectional Appendix it seems they are a lot closer together south of Peterborough, which suggests the limits of what can be supplied can be reached and therefore more sections are required to satisfy the power demand.

What is the limiting factor of the OLE power? Is it the substation and therefore substation upgrades allow greater capacity to the section. Or can the maximum current of the OLE by reached quite easily in busy areas, meaning the only way to increase capacity is to split it into a greater number of sections?

There are a few factors, in order of significance, affecting the two main limits (pantograph voltage and supply capacity):
1. Distance between supplies. There is an exponential worsening of strength with increasing distance from the supply
2. Parallel or island feeding. Bit of a moot point in the UK as our 25kV network is currently island fed throughout.
3. Capacity of supply. Note that the headline figure is pretty meaningless, as it is frequently only the rating of the transformer, which will rarely be usable due to voltage regulation and/or power quality, particularly for larger transformers.
4. Impedance of the OLE (AT, classic, boostered). This mainly affects voltage regulation, it is extremely rare for thermal capacity to be a limiting factor.

So it is impossible to say one supply type is better than another, unless all the other factors are the same; which they generally aren't!

 

[ryan125hst]

I'd be interested to know if trying to stagger train movements for the reason of avoiding a trip is ever actualised in disruption recovery. I'd imagine not, quite enough things to be thinking about. There are also seemingly weak spots that tend to fail almost predictably. The very south of the ECML between Alexandra Palace and Kings Cross tends to get overloaded (notably recently) when lots of trains start moving after there's been disruption.

Yeah I'd be interested to know how much they have to think about it during disruption when lots of trains end up stopped one signal behind the other. When did an overload happen between Alexandra Palace and Kings Cross recently? You'd have thought this wouldn't happen following the upgrades.

That being said, I believe power supplies can be expected to provide over their rated capacity for at least a short time, and (as previously said) there is supposed to be headroom for inconveniently timed loads and to cover for a supply loss. The 'average' feeder station has two grid supplies for each direction away from the site's neutral section. For a 132kV connection it seems to be in the region of 18 miles from my observations. But each of those supplies can be extended to cover the other one's load if that other one goes out (N-1 feeding)and support the timetable. So you could say that each of those supplies is only doing 50% ish of what it could theoretically handle - but actually it's at max capacity because of the requirement to cover for its neighbour.

That makes sense, you need some resilience in the system so one failure doesn't bring all the trains to a stand.

This is at least a little because you can run things a bit closer to the edge and hope a bad set of coincidences don't happen. But I should say, I think this was mainly achieved on ECML by converting from RC/BT feeding to RSC feeding which could be done relatively quickly/easily and removes a non-trivial amount of built in impedance.

What are these types of feeders?

South of Peterborough there are definitely more trains so one would definitely expect a higher density of feeder stations. However, counting neutral sections is not always a reliable way of ascertaining the feeding arrangement as there are often "alternative" midpoints that are normally bridged over but only come into play in certain N-1 situations and there will also be feeder stations without neutral sections if they only have one feed present: there are examples of all of those on ECML

Right I see, so it's not quite as simple as I thought it was as far as neutral sections is concerned. Is there any information anywhere about how it all works?

The potential future loading from electric freight is far greater bearing in mind that we expect conversion from diesel on main lines and a 75% increase in volume by 2050. The new Class 99s can operate in multiple with a combined 'power' of 16,540hp apparently. I appreciate that power figures for locomotives are something of a black art but simplistically that is equivalent to 12,334kW. Imagine a couple of double headers re-starting from rest with 750m loaded intermodal trains either side of Stoke Tunnel one night after a stop for line examination or something. Could be getting on for 50MW for a short period.

That's a huge amount of power. These new electric locos are going to bring challenges due to their huge power output.

Electrification projects not putting in enough grid feeders is a large part of the problem. They are very expensive, so if you can leech off an existing one that tends to happen. It kicks the can down the road for another project basically.

And probably meaning it'll cost a lot more to sort the problem out a few decades down the line!

99s are effectively adhesion limited up to 30mph which limits current draw.

The default maximum current draw per train on NR is 300A which is nominally 7.5MW at 25kV and less if the line voltage drops.
A 99 can draw ~7.2MW at the pantograph and all the Stadler multl-mode locos (88, 93 and 99) have:
i) current limiting to 300A for multi-working hence 2x99 will draw not much more more current than just 1 Cl 99 at medium and higher speeds with the current being equally split between the two locomotives.
ii) tractive effort limiting to ensure the applied TE is not greater that the coupler rating between the second loco and the 1st wagon there is no coupler snap, this would also keep the power draw at low speeds low for two locomotives.

Hence a pair of 99s is only going to take nominally 7.5MW not the 12.34MW you mention which is based on power at the wheel not total power (there has been some rounding and now using enough significant figure in the conversion factor when someone has converted the original kW figures to hp figure you quote hence it doesn't come back to the original kW number up on reconversion)

Interesting. It doesn't surprise me given how large the power output of the Class 99 is. They have multi working capability but it makes me wonder what the benefits would be. I guess traction over double the number of axles helping with adhesion? I hadn't thought about the loads on the coupling either - has that ever been a problem before?

From memory HS1 and the under construction HS2 have higher limits, some of the latest TGV (dasye) are 9280kW, so when operating in pairs could draw nearer 18.5MW

I see the Class 373's have a total output of 12.24MW. As far as I understand, the North of London sets have the same output as the Three Capitals sets. GNER used them in the early to mid 2000's and there are posts on this forum explaining they could accelerate a Class 91 up Stoke Bank on one power car! Were they power limited at all when both power cars were operating or does the distance between pantographs mean the full power can be available and it is the close proximity between the two Class 99 locos that requires the 7.5MW limit?

2MW at stationary / low speeds to avoid welding the graphite composite strip to the wire

I presume this, or lower, applies to the rest of the country. This will cap the charging speed of the tri-mode locos and units charging off the OLE when sat in a platform or yard.

== Doublepost prevention - post automatically merged: 15 Aug 2025 ==

Some more numbers I've come across
- average rating of a 132/25kV supply is between 25-26.5 MVA, but can be more like 18MVA on the low end
- supergrid connections whether to 25kV or 50kV is between 80 and 104MVA

Yes I found an old post earlier suggesting that Retford's feeder is only 18MVA, so not a huge amount at all. Would more than this be available in a normal situation thanks to feeders further up the line? A 5 car Azuma has a total power in electric mode of 2,712kW, so 5,424kW for a pair, with a 9 car Azuma coming in at 4,520kW. So three 9 or 10 car Azumas accelerating hard in the section would take a large percentage of this capacity, and let's not forget each train may be using 300kW or more of auxiliary power for the air conditioning and catering etc.

 

[Bald Rick]

When did an overload happen between Alexandra Palace and Kings Cross recently? You'd have thought this wouldn't happen following the upgrades.

Ferme Park feeder has been offline for a while.

 

[GraphitProjekt]

Yeah I'd be interested to know how much they have to think about it during disruption when lots of trains end up stopped one signal behind the other. When did an overload happen between Alexandra Palace and Kings Cross recently?

27th July. https://www.railforums.co.uk/thread...stevenage-27-07-25.290023/page-3#post-7403011. It followed disruption that meant lots of trains were suddenly accelerating out of the depot(s) around Alexandra Palace. And there was another failure in the area on 1st August but I'm not sure if that was power supply or wires coming down.

You'd have thought this wouldn't happen following the upgrades.

You would think yeah! ECML PSU phase 1 formally covered Wood Green to Bawtry so the southern extent is Wood Green feeder station (at Alexandra Palace). Wood Green FS was upgraded but it's not clear to me exactly how, but it involved building new substation a few hundred feet down the line. Maybe it has a higher rated transformer or switchgear I'm not really sure. But according to that wording, none of the substations south of Wood Green were touched by the PSU which includes Ferme Park FS which I believe is responsible for supplying the line from Alexandra Palace south all the way to Kings Cross. Both Wood Green and Ferme Park are supplied off the same 132kV (underground) line between Tottenham 275/132 and Hornsey 132/33 so in effect are kinda one feeder station. However...

Ferme Park feeder has been offline for a while.

This explains things. The London end of ECML and Hertford Loop are basically in their N-1 situation day-in day-out. I would assume this means the line between Kings Cross/Drayton Park/Canal Jnc through Wood Green up to Hertford North is all coming off the single Wood Green supply. It could also be the case that just Kings Cross-Wood Green is being supplied by Wood Green and Corey's Mill (whose feed normally ends at Hertford North) is powering the whole Hertford Loop to Bowes Park neutral section. The normal arrangement is like this. Although this is all my own educated guessing so it could be wrong.

That makes sense, you need some resilience in the system so one failure doesn't bring all the trains to a stand.

With two failures, so a whole dual substation gone out (aka N-2), the next door feeders can extend beyond mid point substations but under N-2 there is expected to be timetable and likely top speed alterations. I was once watching a cab ride in third rail lane (I think around Seaford) and a note came up on screen that owing to a power failure drivers were instructed not to go beyond notch 2 on the power controller in a particular area.

What are these types of feeders?

These are essentially methods of protecting lineside twisted pair copper cables for signalling and telecoms (and to an extent protecting railway neighbours) from EMI made by the OLE. RC/BT means an arrangement with "return conductor" and "booster transformer". The return conductor is the thicker wire hung on the outside of OLE structures and is connected to the running rails at regular intervals. It's meant to collect the return portion of the circuit and run it back to the feeder station adjacent to the physical location of the live catenary wires. Since the current is travelling in the opposite direction to that in the live wires, it's meant to counteract the EMI.Booster transformers are an addition to having a return conductor and uses power from the OLE circuit to "force" more opposite negative current to flow through the return conductor thus improving the overall EMI situation - but it comes with a cost of higher impedance in the OLE system.

The proper description of a BT is:

"A BT is a 1:1 ratio current transformer, with the primary winding connecting in series with the OLE so that traction current is routed through it. The secondary winding of the BT is connected in series across an electrical break in the RC. The current in the primary induces an equal and opposite current in the secondary winding, and so drives current in the RC; and this current can only come from the rail at a bond connection midway between BTs called the midpoint connection."

Since you said you actually are an electrical engineer this will make more sense to you than it does to me to be honest. This description is from Garry Keenor's book on OLE which freely available and a really good detailed overview of the whole situation.

RSC means Return Screen Conductor. This method does away with BTs and return conductors, and instead runs thick insulated cable within the lineside trunking (where the telecoms/signal cables that need to protected are) that, I guess, take the hit of EMI instead of inducing currents in the S&T wires.

== Doublepost prevention - post automatically merged: 16 Aug 2025 ==

There isn't much RC/BT left on ECML - I think just Hertford Loop and Leeds branch. But it's quite common on all electrification done between about 1960 and 2010 that hasn't been converted to something else. So it's all over the place in Anglia Region and the Birmingham area (not WCML in general though).

There is also 2x 25kV Autotransformer (AT) feeding which is where another live conductor at -25kV is hung up in a similar place to an RC called an AT feeder (ATF). The ATF connects every ~10km to the +25kV catenary through one or more autotransformers that are somehow able to convert the negative current in the ATF into more current for the contact wire. This is done at a kind of minor substation called ATS (auto-transformer site) every ~10km or through ATs fitted to other railway substations (mid-points, sectioning substations, feeder stations). Because the current in the ATF wire is opposite to that in the catenary, the AT system effectively has EMI mitigation sort of built in. In practice RSCs are always installed.

In effect the railway is fed at 50kV allowing in theory for more power and/or a greater distance between grid supply points. Auto-transformer feeder stations (ATFS) are only ever fed by 275-400kV lines to a 50kV transformer that is centre tapped to the 0V running rails, the +25kV end goes to the contact wire and the -25kV side goes to the ATF. Because of this it's often called 2x 25kV or 25-0-25kV. The actual technical detail of how ATs convert from -25kV to +25kV and all the directions currents are going in to reach the train is well beyond me and I've never quite been able to get my head around the description. This (left) is what it looks like though, note the outside wire with a full 25kV insulator. Whereas with RC/BT (to the middle) the outside wire is much thicker and not insulated (very much) as it's all return. And finally (right), that's what a BT looks like

Right I see, so it's not quite as simple as I thought it was as far as neutral sections is concerned. Is there any information anywhere about how it all works?

Not really, definitely not all in one place but there are clues scattered around. It's sort of become my main hobby since about October last year to try and figure out the arrangement of grid supply points and feeding boundaries as much as i can mostly from collecting partially useful sources cross-referencing them together often requiring some healthy guesswork. Unlike signalling, there is not as much enthusiast interest in establishing the lay of the land of this topic.

I presume this, or lower, applies to the rest of the country. This will cap the charging speed of the tri-mode locos and units charging off the OLE when sat in a platform or yard.

OLE as a system is much more capable of charging in motion than charging statically especially when compared to third rail (or third rail like stuff - as on Greenford branch)

== Doublepost prevention - post automatically merged: 15 Aug 2025 ==


Yes I found an old post earlier suggesting that Retford's feeder is only 18MVA, so not a huge amount at all.

I'm pretty sure Retford FS is getting or has been upgraded to have a more powerful connection, I'm not sure what it is however. A higher power connection at Retford was agreed with the DNO many years ago but then retrospectively rejected by DNO some time later. This was enough of a big deal to make the press. (https://www.modernrailways.com/article/hardware-hamstrings-ecml-2022-timetable-plans)

As part of this project, Network Rail had applied for a firm service capacity uplift from WPD at its existing feeder station at Retford. According to NR, this would ‘essentially’ increase the power draw to the rated capacity of the feeder station equipment. However, WPD [now NGED] highlighted existing voltage imbalance issues, known as negative phase sequence non compliances, both at Retford and Grantham feeder stations. These would have to be resolved before any increase would be granted. This imbalance results from electric traction power supplies drawing a single phase from the three-phase grid...According to Network Rail, a technical solution will take ‘two to four years’ to deliver. This is likely to be the installation of a Static Frequency Converter (SFC).

That was in 2021, I haven't found anything yet on what they did about it but the clearly worked something out because the replacement to Grantham did come online only this year, although I haven't seen anywhere that the new Retford feeder is operating yet.

Would more than this be available in a normal situation thanks to feeders further up the line?

The N-2 for Retford is probably a lot better than it used to be with new feeders either side (Grantham new and Potteric Carr) but FSs can't work together so in normal operation it's just a bottle-neck. Technically it is theoretically possible to run Potteric Carr in parallel with one other substation since Potteric Carr is an SFC but this capability has been explicitly stated as not an option for now

A 5 car Azuma has a total power in electric mode of 2,712kW, so 5,424kW for a pair, with a 9 car Azuma coming in at 4,520kW. So three 9 or 10 car Azumas accelerating hard in the section would take a large percentage of this capacity, and let's not forget each train may be using 300kW or more of auxiliary power for the air conditioning and catering etc.

 

[Richard123]

Yeah I'd be interested to know how much they have to think about it during disruption when lots of trains end up stopped one signal behind the other. When did an overload happen between Alexandra Palace and Kings Cross recently? You'd have thought this wouldn't happen following the upgrades.


That makes sense, you need some resilience in the system so one failure doesn't bring all the trains to a stand.


What are these types of feeders?


Right I see, so it's not quite as simple as I thought it was as far as neutral sections is concerned. Is there any information anywhere about how it all works?


That's a huge amount of power. These new electric locos are going to bring challenges due to their huge power output.


And probably meaning it'll cost a lot more to sort the problem out a few decades down the line!


Interesting. It doesn't surprise me given how large the power output of the Class 99 is. They have multi working capability but it makes me wonder what the benefits would be. I guess traction over double the number of axles helping with adhesion? I hadn't thought about the loads on the coupling either - has that ever been a problem before?


I see the Class 373's have a total output of 12.24MW. As far as I understand, the North of London sets have the same output as the Three Capitals sets. GNER used them in the early to mid 2000's and there are posts on this forum explaining they could accelerate a Class 91 up Stoke Bank on one power car! Were they power limited at all when both power cars were operating or does the distance between pantographs mean the full power can be available and it is the close proximity between the two Class 99 locos that requires the 7.5MW limit?


I presume this, or lower, applies to the rest of the country. This will cap the charging speed of the tri-mode locos and units charging off the OLE when sat in a platform or yard.

== Doublepost prevention - post automatically merged: 15 Aug 2025 ==


Yes I found an old post earlier suggesting that Retford's feeder is only 18MVA, so not a huge amount at all. Would more than this be available in a normal situation thanks to feeders further up the line? A 5 car Azuma has a total power in electric mode of 2,712kW, so 5,424kW for a pair, with a 9 car Azuma coming in at 4,520kW. So three 9 or 10 car Azumas accelerating hard in the section would take a large percentage of this capacity, and let's not forget each train may be using 300kW or more of auxiliary power for the air conditioning and catering etc.

The project upgrade was from Wood Green north, there was a PWI video a few years ago that explained. It was originally scoped as an AT conversion, and building AT through a narrow corridor is near impossible. See also WCML.

However, no electrification can handle a queue of trains starting simultaneously, which becomes more of an issue with ETCS (long block sections tend to naturally impose some stagger). There are procedures to stagger restarts if needed.

 

[D365]

… which becomes more of an issue with ETCS (long block sections tend to naturally impose some stagger). There are procedures to stagger restarts if needed.

But on the flip side, ETCS ought to make it easier to manage each individual train in such a scenario.

 

[Dazza12]

Hi,

I don't even know how I cocked that one up! Guess I'll have to retake primary school maths at some point lol

Some more numbers I've come across
- average rating of a 132/25kV supply is between 25-26.5 MVA, but can be more like 18MVA on the low end
- supergrid connections whether to 25kV or 50kV is between 80 and 104MVA

Note that the 18MVA and 26.5MVA transformers are largely identical and will therefore have the same short term (i.e. say 2 or 3 minute) ratings. The only difference is that the 26.5MVA ones have fans on their radiators which improve their long term ratings.

 

[hwl]

Interesting. It doesn't surprise me given how large the power output of the Class 99 is. They have multi working capability but it makes me wonder what the benefits would be.

Multi working on diesel as they have noticeable less TE than 66 if there is just one of them

I guess traction over double the number of axles helping with adhesion?

At low - mid speeds

I hadn't thought about the loads on the coupling either - has that ever been a problem before?

Lots of rules and calculations in place to prevent it being a problem.

I presume this, or lower, applies to the rest of the country. This will cap the charging speed of the tri-mode locos and units charging off the OLE when sat in a platform or yard.

Not too much in practice.

 

[Dr Hoo]

I'm still surprised that more people don't seem to have much ambition when it comes to getting the best out of Class 99s (and similar modern electric/multimode freight traction). Over 70 years ago the Woodhead 1,500V dc electrification was commissioned with a basic method of working from Wath up to Penistone with two class EM1 locomotives in tandem at the front and two more in tandem banking at the rear. (Multiple working fitting came later.) All four locomotives working with both pantographs raised.

People are always raising the issue of slow freights struggling up Shap (and other inclines); the need to raise the maximum speed for intermodals to 90mph; the need for automatic couplings; and the scope for ERTMS/ETCS to allow for services to follow more closely. It is also said that with the HS2 Scottish trains it will no longer to have the passenger timetable 'bunched' north of Carnforth/Oxenholme. Similar comments apply in the opposite direction.

My vision is that the freight loops, wherever provided, should be reconfigured with greater length and much higher speed paintwork. So, if a 125mph Pendolino passes (say) Grayrigg the points can switch the moment that it passes, movement authority issued to a looped freight with 2 x Class 99 train locomotives and 2 x Class 99 bankers so that it can make a 'dragster start' with all eight pantographs raised. Six miles further on, at Tebay, a similar consist is let out behind the Pendolino around three minutes later, also with a dragster start. The two freights can continue at full thrash until they have achieved 90mph (subject, obviously, to any local PSRs).

Clearly such methods of operation would require a new approach to power feeding. However, the 15Es in South Africa haul trains over 40,000 tonnes uphill with 5 x 4,500kW locomotives in radio distributed power 'multiple' operation and have done for 15 years. (Yes, I know that it's 50kV, not 25kV, and narrow gauge but fundamentally I don't see that I'm expecting anything really novel.)

 

[bahnause]

My vision is that the freight loops, wherever provided, should be reconfigured with greater length and much higher speed paintwork. So, if a 125mph Pendolino passes (say) Grayrigg the points can switch the moment that it passes, movement authority issued to a looped freight with 2 x Class 99 train locomotives and 2 x Class 99 bankers so that it can make a 'dragster start' with all eight pantographs raised. Six miles further on, at Tebay, a similar consist is let out behind the Pendolino around three minutes later, also with a dragster start. The two freights can continue at full thrash until they have achieved 90mph (subject, obviously, to any local PSRs).

The selling point of the Class 99 is that it can transport large loads with few operating resources (=locos) and therefore cost-effectively. Providing four locomotives per train only needed in certain places does not quite fit in with this concept.

 

[edwin_m]

Looking at the relative timings, just getting rid of diesel freight on Shap and Beattock would bring a lot of benefit. Going from one electric to two or four is diminishing returns.

 

[Dr Hoo]

The selling point of the Class 99 is that it can transport large loads with few operating resources (=locos) and therefore cost-effectively. Providing four locomotives per train only needed in certain places does not quite fit in with this concept.

I thought that the use of the term ‘banking’ locomotives would make clear that they wouldn’t run all the way from Felixstowe to Coatbridge or wherever, rather be confined to applicable areas with steep gradients, e.g. loosely Carnforth-Penrith/Carlisle. Given that it would be GBR that gains the benefit of capacity uplift there is nothing to stop GBR (or its contractor) providing banking services to all operators.

We are supposed to be achieving a 75% increase in freight by 2050 at the same time as converting existing diesel flows under the wires to electric traction so we’re going to need some pretty radical thinking and changes.

 

[MarkyT]

I see the Class 373's have a total output of 12.24MW. As far as I understand, the North of London sets have the same output as the Three Capitals sets. GNER used them in the early to mid 2000's and there are posts on this forum explaining they could accelerate a Class 91 up Stoke Bank on one power car!

NoL sets found work on the continent after they were surplus to requirements here. The tranche of Three Capitals sets released when the Velaros arrived seemed to raise no interest from second operators on withdrawal and have been largely scrapped, or have become spares christmas trees at depots. Perhaps they were unattractive to operators due to their huge inflexible length and more sluggish acceleration? I think the NoL sets are a near equivalent in power/weight to the first-gen LGV Sud-Est sets, so they could probably keep up on that busy, steeply graded route to Lyon, while the 3Cs likely could not.

 

[bahnause]

I thought that the use of the term ‘banking’ locomotives would make clear that they wouldn’t run all the way from Felixstowe to Coatbridge or wherever, rather be confined to applicable areas with steep gradients, e.g. loosely Carnforth-Penrith/Carlisle. Given that it would be GBR that gains the benefit of capacity uplift there is nothing to stop GBR (or its contractor) providing banking services to all operators.

This also requires additional locomotives. With presumably very inefficient diagrams and high kilometre costs.

We are supposed to be achieving a 75% increase in freight by 2050 at the same time as converting existing diesel flows under the wires to electric traction so we’re going to need some pretty radical thinking and changes.

Making production more expensive will not achieve this goal. In the Alpine countries, a lot of money is being invested in base tunnels to prevent expensive solutions such as pushing locomotives and double heading.

 

[Dr Hoo]

This also requires additional locomotives. With presumably very inefficient diagrams and high kilometre costs.

Yes, but so does running a greater number of smaller trains to shift the same tonnage, and consumes more scarce capacity into the bargain.

Making production more expensive will not achieve this goal. In the Alpine countries, a lot of money is being invested in base tunnels to prevent expensive solutions such as pushing locomotives and double heading.

Fair enough. I can see the benefits of long base tunnels under Shap, Beattock, the Peak District, Pennines, Dartmoor and so on. However, until they are built…

 

[edwin_m]

Attaching and detaching bankers will cost more time than is saved by running faster up the incline.