Power drain at terminal stations

[Railcar]

A train moving off must consume a lot of power. What if several trains are on the move together? The load on the station's traction power supply must be a big peak. Are the timetables aranged so that there is an timed offset to the number of trains that can be starting at one time? Are signallers trained to avoid simultanious Platform Starter greens?
These questions were prompted by waiting for my train at London Bridge, where all 15 platforms can be occupied by trains drawing current on the 750v d.c. supply (OK, except for the Uckfield trains)

 

[The Planner]

A train moving off must consume a lot of power. What if several trains are on the move together? The load on the station's traction power supply must be a big peak. Are the timetables aranged so that there is an timed offset to the number of trains that can be starting at one time? Are signallers trained to avoid simultanious Platform Starter greens?
These questions were prompted by waiting for my train at London Bridge, where all 15 platforms can be occupied by trains drawing current on the 750v d.c. supply (OK, except for the Uckfield trains)

It can be modelled yes, it tends to be the amount of trains per hour in a section that is raised as an issue. I am not aware of any restrictions on trains starting at the same time timetable wise, I am trying to think of locations where you would have multiple trains doing that, it would be pure luck in a lot of cases if there were several at the same time. This sort of issue would normally flag congested infrastructure and force a solution like various sections of the network currently have.

 

[Peter Sarf]

Bear in mind the maximum number of trains moving simultaneously at most termini will be limited by the number of tracks feeding the station and not the number of platforms. For London Bridge it is 7 trains on the through platforms (platforms 1-3 Cannon Street, 4-5 ThamesLink, 6-9 Charing Cross reducing to 2 tracks) maximum movements 7. For the Termini part (6 platforms) there might be six tracks feeding it (3 in and 3 out). So total 10 trains departing is possible.

Somewhere like Victoria has 19 platforms iirc. Probably 8 tracks but half in so 4 departing simultaneously. Not so heavy BUT up a steep gradient to get over the Thames.

Compare with a through station like East Croydon. Six platforms so six trains departing simultaneously possible - more than Victoria !. CLapham Junction with 17 platforms - platform 7&8 share one line in and out, platform 17 conflicts with departures from Platform 15 BUT still 15 trains possible !.

So don't worry about termini !.

 

[Bald Rick]

A train moving off must consume a lot of power. What if several trains are on the move together? The load on the station's traction power supply must be a big peak. Are the timetables aranged so that there is an timed offset to the number of trains that can be starting at one time? Are signallers trained to avoid simultanious Platform Starter greens?
These questions were prompted by waiting for my train at London Bridge, where all 15 platforms can be occupied by trains drawing current on the 750v d.c. supply (OK, except for the Uckfield trains)

Timetables are not arranged for this. The power supply is sized accordingly. For obvious reasons, you couldn’t have a train departing from every platfrom at London Bridge concurrently.

Besides, the power supply covers a much wider area than just the station. For example the whole South West suburban area as far out as Windsor, Shepperton and Surbiton is all fed from Wimbledon. And again, for obvious reasons, you cant have every train accelerating at the same time.

 

[Deepgreen]

A train moving off must consume a lot of power. What if several trains are on the move together? The load on the station's traction power supply must be a big peak. Are the timetables aranged so that there is an timed offset to the number of trains that can be starting at one time? Are signallers trained to avoid simultanious Platform Starter greens?
These questions were prompted by waiting for my train at London Bridge, where all 15 platforms can be occupied by trains drawing current on the 750v d.c. supply (OK, except for the Uckfield trains)

The short answer is; no - timetables are based on route requirements, signal spacing, etc. Even if timetables were based on this, the chances of a few trains being slightly off-time are high, so the planning would be in vain anyway. Supplies are far wider than just the local terminal area in any case. However, some routes' sections do have limits on movements owing to supply limitations - e.g. Bournemouth to Weymouth.

 

[Dr Hoo]

Although not strictly a 'terminal station' in the sense raised by the OP, this could be a problem at some depot locations. When I worked on the South Eastern Division (nearly 40 years ago) Ramsgate depot (and station) would be full overnight with many trains 'cut in' during cleaning or when about to be shunted. Besides this static load several moves could be made simultaneously on different parts of the layout, e.g. to or from the carriage washer. Two moves with 12 cars were fine but a third move could pull the breakers out.

 

[contrex]

When the Bournemouth electrification scheme was new, I read there was a problem with breakers coming out if too many simultaneous, or nearly so, starts were made. Upgrades were ordered. Wasn't a new beefier substation put in for Blackfriars as part of the Thameslink upgrade?

the whole South West suburban area as far out as Windsor, Shepperton and Surbiton is all fed from Wimbledon.

Controlled, but not 'fed', surely?

 

[swt_passenger]

Controlled, but not 'fed', surely?

AIUI Wimbledon is where the grid supply connection to the railway is, it’s then distributed trackside to substations all around the Wessex inner area, and they are all connected in parallel on the DC side via trackside equipment. The electrical control room for Wessex inner is at Raynes Park, I believe.

 

[Bald Rick]

Wasn't a new beefier substation put in for Blackfriars as part of the Thameslink upgrade?

Its a substation, yes, but this is for distribution. it’s not a Connection to the grid.

Controlled, but not 'fed', surely?

AIUI Wimbledon is where the grid supply connection to the railway is, it’s then distributed trackside to substations all around the Wessex inner area, and they are all connected in parallel on the DC side via trackside equipment. The electrical control room for Wessex inner is at Raynes Park, I believe.

Is the right answer!

 

[contrex]

You don't mean, I take it, that all the juice rails in the Wessex area are in parallel with each other? Wouldn't that make the fault current in the event of a short, gigantic?

 

[apk55]

I would have thought the power supply for a complex station area would be divided up into many different sections each with its own circuit breaker which would be set at level of maybe 2 train loads max. This would be done so that if there is a short circuit fault it can be detected and cleared - if the whole station was one circuit it would difficult to detect a fault and the potential fault current would be massive possibly beyond the capacity of the train mounted circuit breakers.
It also may be easier to use several smaller substations than one big one again for fault current reasons as well as using a standard design.

Do not forget forget substation can normal tolerate a short term overload of 2 to 300% for a couple of minuets.
Also modern trains when accelerating can only take maximum line current when at speed (typically above 40MPH) - at starting and low speed the current would be much lower even when on full acceleration.

It must be quite a design exercise to design the power supply system for a complex station investigating all the fault possibilities.

 

[Snow1964]

In past have been approaching Waterloo in morning rush hour (pre 2017 alterations), 5 trains drew up awaiting platforms, 3 left simultaneously and once clear, all 5 waiting for platforms started moving together side by side.

Although speeds were low, 8 trains accelerating together must be hefty load, but probably only lasted under a minute before they were coasting and substation got too warm.

 

[edwin_m]

You don't mean, I take it, that all the juice rails in the Wessex area are in parallel with each other? Wouldn't that make the fault current in the event of a short, gigantic?

That's exactly what happens, and getting DC substations to trip on short circuit but not when several trains are motoring simultaneously is a bit of a black art. That's why we have fires from things like drinks cans wedged against the third rail, and there were considerably more serious fires in the past when protection wasn't as good.

In practice the resistance of the third rail means that only the nearest few substations have any effect at a particular place, and when a substation trips out various breakers will disconnect so no other substation is feeding that section.

 

[InTheEastMids]

Although speeds were low, 8 trains accelerating together must be hefty load, but probably only lasted under a minute before they were coasting and substation got too warm.

Yes, remembering that from the technical perspective, this type of electrical equipment usually has short-term and continuous ratings.

From a commercial perspective there will be capacity agreements that limit what the railway should draw from each supply point (financial penalties for going over). Generally, a commercial customer will get a capacity agreement in line with their demand, rather than the technical limitations imposed by the capacity of transformers and cables - e.g. if a factory replaces one machine with a new energy efficient machine (or replaces it with a machine in a faraway country with lax labour and environmental laws), then one of the savings is a reduction in cost of capacity for electricity.

This approach also means that if you add a new demand like EV charging it can sometimes be done with a stroke of a pen, or very minor changes to circuit breakers etc. This is increasingly rare however, as the proliferation of renewable generation and EV charging means a lot of this surplus has been used up. As the railway has become more intensively used by longer, heavier, higher-performing trains, then that again is about expensive new grid capacity.

Back to a point the original poster asked about

What if several trains are on the move together? The load on the station's traction power supply must be a big peak.

For the various reasons supplied by others, having a lot of trains diversifies the load and smooths this out and the demand becomes more predictable percentage of the theoretical max. You could see something similar if you look at your own home's demand for electricity, which would appear quite spiky as you turned on things like kettles and toasters versus what it looks like if you aggregate many houses together where you just kind of get very smooth rising and falling demand. At the other end of the scale, if you imagine a very rural bit of electrification, you might size it on the basis of only 1 train being in that section at a time. The train leaves and demand goes almost instantly from a low level to around 100% of the theoretical demand you've sized for, which might be more of an issue, especially if you then decide to replace something like a Cl 319 (~1 MW) with a 350 (~2 MW), or double the train length etc.

 

[Bald Rick]

In past have been approaching Waterloo in morning rush hour (pre 2017 alterations), 5 trains drew up awaiting platforms, 3 left simultaneously and once clear, all 5 waiting for platforms started moving together side by side.

Although speeds were low, 8 trains accelerating together must be hefty load, but probably only lasted under a minute before they were coasting and substation got too warm.

If there were 5 approaching, it was on the immediate approaches, and therefore acceleration (to 15/20mph) would have been a matter of secinds, and unlikely to use full power.

 

[boiledbeans2]

Also modern trains when accelerating can only take maximum line current when at speed (typically above 40MPH) - at starting and low speed the current would be much lower even when on full acceleration.

Is the answer.

Mechanical power = force x velocity.* When pulling away from stations, the speed is low, and therefore, power consumption is low.

*There will also be electrical losses to calculate the electrical power. At low speeds, the powertrain would be roughly 50-60% efficient.

 

[Tunnel Bore]

In past have been approaching Waterloo in morning rush hour (pre 2017 alterations), 5 trains drew up awaiting platforms, 3 left simultaneously and once clear, all 5 waiting for platforms started moving together side by side.

Although speeds were low, 8 trains accelerating together must be hefty load, but probably only lasted under a minute before they were coasting and substation got too warm.

We need to remember the difference between power and tractive effort (or torque). While accelerating there is a large torque developed by the traction motors and large current in the motors but that is fed via some power electronics (inverter etc) which draws a different current from the supply that is dictated by the power being delivered. If the speed is low then the power is low because power = speed x torque. The current drawn from the supply is then current = power / voltage.

Hill-climbing or accelerating at high speed requires a high torque at a high speed and therefore a high power and high current from the supply.
It was rather different in the days before inverters and other power electronics. Series DC traction motors connected via starting resistors would draw very large currents direct from the supply at low speed while starting and develop extremely high torque.

 

[hwl]

If there were 5 approaching, it was on the immediate approaches, and therefore acceleration (to 15/20mph) would have been a matter of seconds, and unlikely to use full power.

(D)EMUs are effectively adhesion limited up to ~20mph and the software limits the traction motor power to just slightly more than that needed to deliver the TE at a given speed so the onboard electrical equipment doesn't get too warm. So traction power (ignoring aux.,. and hotel loads) is increases effectively linearly between 0 and 20mph. The fun starts when you have lots of trains doing 20+mph...
Motor efficiency is pretty grim below 30% of max train speed (hence different fixed caps than from 0-20mph) and peak motor efficiency and torque is around 60-70% of max train speed.

 

[Nicholas Lewis]

We need to remember the difference between power and tractive effort (or torque). While accelerating there is a large torque developed by the traction motors and large current in the motors but that is fed via some power electronics (inverter etc) which draws a different current from the supply that is dictated by the power being delivered. If the speed is low then the power is low because power = speed x torque. The current drawn from the supply is then current = power / voltage.

The other advantage of modern traction electronics is that the inverters will control the maximum current draw as well.

Hill-climbing or accelerating at high speed requires a high torque at a high speed and therefore a high power and high current from the supply.
It was rather different in the days before inverters and other power electronics. Series DC traction motors connected via starting resistors would draw very large currents direct from the supply at low speed while starting and develop extremely high torque.

They would although worse case on DC was when the motors went into weak field (35-40mph) as that would impose a hefty spike on the traction system and in the event that a 12car (3x4) were pretty well matched that was often what tripped a CB especially where that happened near a substation

 

[paul1609]

Although not strictly a 'terminal station' in the sense raised by the OP, this could be a problem at some depot locations. When I worked on the South Eastern Division (nearly 40 years ago) Ramsgate depot (and station) would be full overnight with many trains 'cut in' during cleaning or when about to be shunted. Besides this static load several moves could be made simultaneously on different parts of the layout, e.g. to or from the carriage washer. Two moves with 12 cars were fine but a third move could pull the breakers out.

Until the Electrostars were introduced the Littlehampton Branch and Littlehampton Depot were fed from a substation at Arundel Junction. When the 377s were introduced a new substation had to be built at Littlehampton because the existing arrangement couldnt handle the load of 4 x 12 car 377s in the depot heating plus 2 in the platforms with one of them moving.

 

[edwin_m]

Ian Walmsley mentioned in December's Modern Railways that if there are problems (presumably those resulting in multiple trams at a stand nose to tail), controllers have to instruct each tram to re-start in sequence, otherwise several drawing current simultaneously could trip out the supply.

 

[Nicholas Lewis]

Ian Walmsley mentioned in December's Modern Railways that if there are problems (presumably those resulting in multiple trams at a stand nose to tail), controllers have to instruct each tram to re-start in sequence, otherwise several drawing current simultaneously could trip out the supply.

Thats always been an issue with trams from day one as unlike trains they can all move simultaneously but even with ATO there is still a slight lag between trains starting to move if they all get bought to a stand

 

[edwin_m]

Thats always been an issue with trams from day one as unlike trains they can all move simultaneously but even with ATO there is still a slight lag between trains starting to move if they all get bought to a stand

Indeed, if there is a queue of trains at successive signals the second one can't start to move until the first has passed the signal and cleared the overlap. However, a modern EMU draws more current as speed increases, and with greater weight and better performance that EMU will probably be drawing more current than a traditional DC-motored one. That's the sort of reason why the software in third rail EMUs is set up to limit current draw to being similar to traditional EMUs at the same speed.

 

[Tim M]

What seems to be missing from this thread is the effect of regenerative braking from approaching trains pushing power back into the traction system to supplement power from the grid. This can never be 100% of course, the rules of physics leading to brake fade necessitating blended changeover from regenerative to friction braking.

From memory of a new metro in China, regenerative brake fade occurred at about 12km/h after which friction braking automatically cut in. The Chinese hoped for a lower figure of about 9km/h but the Automatic Train Operation system couldn't cope and trains overshot platforms as a result.

I‘m not up to date with how many trains in the U.K. are equipped with regenerative braking capabilities, or for that matter if substations are capable of managing power return from the third rail. Maybe others will know more.

 

[swt_passenger]

What seems to be missing from this thread is the effect of regenerative braking from approaching trains pushing power back into the traction system to supplement power from the grid. This can never be 100% of course, the rules of physics leading to brake fade necessitating blended changeover from regenerative to friction braking.

From memory of a new metro in China, regenerative brake fade occurred at about 12km/h after which friction braking automatically cut in. The Chinese hoped for a lower figure of about 9km/h but the Automatic Train Operation system couldn't cope and trains overshot platforms as a result.

I‘m not up to date with how many trains in the U.K. are equipped with regenerative braking capabilities, or for that matter if substations are capable of managing power return from the third rail. Maybe others will know more.

Older DC substations aren’t reversible. Regeneration has still been widely used in the third rail area for around 10 years or more, but it relies on the network being able to accept the power, so individual EMUs cease regeneration if line voltage exceeds a set value. This is a direct benefit of everything being connected in parallel over large areas, as has been mentioned.

 

[snowball]

Older DC substations aren’t reversible. Regeneration has still been widely used in the third rail area for around 10 years or more, but it relies on the network being able to accept the power, so individual EMUs cease regeneration if line voltage exceeds a set value. This is a direct benefit of everything being connected in parallel over large areas, as has been mentioned.

Surely if a powering train and a braking train are in the same section then they could help each other out more than ten years ago? I seem to remember hearing about Woodhead, for example?

 

[swt_passenger]

Surely if a powering train and a braking train are in the same section then they could help each other out more than ten years ago? I seem to remember hearing about Woodhead, for example?

Ten years ago in this context refers to modern stock such as SWT Desiros not having the regeneration function enabled until about then. On delivery they left it switched off because AIUI they were being highly cautious, they were worried line voltage might increase too much during normal random events, especially at the extremities, such as at Weymouth.

Another problem back then was that allowing regeneration in specific areas would have affected other systems, eg allowing regen to raise the line voltage towards 900V in the Wimbledon area would have likely blown up older District line trains. The inner area wasn’t even increased to 750V from 660V until comparatively late on, around 20116/17. DC Regen in Wessex was still an ongoing project in the early CP5 enhancement plan updates, part of the work involved separating out the District line power supplies, I think the Waterloo & City had been separated in an earlier phase.

 

[ac6000cw]

Surely if a powering train and a braking train are in the same section then they could help each other out more than ten years ago? I seem to remember hearing about Woodhead, for example?

Yes, the Woodhead scheme supported regeneration, to provide electric braking on the steep downhill sections for heavy freight trains. AFAIK if there were insufficient trains in the area to make use of the regen power, the excess got dissipated (as heat to the atmosphere) in resistor banks at the feeder stations. The basic design of the scheme dated back to the LNER pre-WW2, so over 80 years ago...

 

[edwin_m]

Older DC substations aren’t reversible. Regeneration has still been widely used in the third rail area for around 10 years or more, but it relies on the network being able to accept the power, so individual EMUs cease regeneration if line voltage exceeds a set value. This is a direct benefit of everything being connected in parallel over large areas, as has been mentioned.

For third rail, not as much as might first appear. As the distance between the braking train and the nearest motoring train increases, so too does the voltage rise (the reverse of a voltage drop) created by the current flowing from one to the other. Once the voltage at the regenerating train reaches the 900V maximum, it has to limit or stop regeneration. This is much less of a problem on 25kV because the currents are so much lower - another reason why lower voltages are better for intensive suburban railways when there's likely to be another train motoring not far away, than for longer-distance routes.

 

[DerekC]

Ten years ago in this context refers to modern stock such as SWT Desiros not having the regeneration function enabled until about then. On delivery they left it switched off because AIUI they were being highly cautious, they were worried line voltage might increase too much during normal random events, especially at the extremities, such as at Weymouth.

Another problem back then was that allowing regeneration in specific areas would have affected other systems, eg allowing regen to raise the line voltage towards 900V in the Wimbledon area would have likely blown up older District line trains. The inner area wasn’t even increased to 750V from 660V until comparatively late on, around 20116/17. DC Regen in Wessex was still an ongoing project in the early CP5 enhancement plan updates, part of the work involved separating out the District line power supplies, I think the Waterloo & City had been separated in an earlier phase.

It wasn't just SWT that was being highly cautious over regen on the DC network - Network Rail had cold feet as well. A certain senior engineering gentleman from the operator community, then on secondment to DfT, did an excellent job of embarrassing the industry into doing something about it in about mid 2007. There was then a long programme of testing, carried out on the Shepperton branch and between Dorking and Horsham, with a full analysis of the DC network by NR to ensure there were no high risk area. I think regen went live somewhere about 2010 on parts of the DC network. You are right about the Wimbledon area and the LU worry about old District line trains, although I seem to remember that to blow one up you would have probably needed every single main line train in the area to be on full regen braking at once. And nobody knew what "blowing up" meant anyway!