Science Articles

Giving Rail the Green Light

Wed, 30th Sep 2015

How our railways are saving energy by braking

Heather Douglas

You might have heard that taking the train is the ‘green’ choice. It is. Rail takes far less energy to transfer a passenger from A to B than air or road. In the European Union, rail transports around 6% of all passenger journeys by distance travelled, but only accounts for 2% of the total transport energy! That is pretty good going, but there’s still plenty of opportunity to make our railways greener.

Running train services in the UK takes the equivalent of 12 Tera Watt hours (TWh) of electrical energy every year. That’s roughly the same as boiling half a million kettles a million times each. In fact, meeting the EU target of a 6% reduction in energy would save enough electricity for every UK resident to have an extra 435 cups of tea annually! Unfortunately though, it isn’t current policy to save energy in order to supply commuters with free cups of tea.

Cup of tea on the train

Where is the energy ‘used’?

Before we think about saving energy, we need to know where it is ‘used’. Basic physics dictates that energy cannot be created or destroyed, but is transformed from one form to another. Energy is considered ‘used’ once it is converted into a non-reusable form – such as heat, light, noise or vibration.

Trains themselves use energy in lots of ways: to supply heating, lighting and cooling on board; to overcome the aerodynamic resistance to motion; and in less than perfect power electronics in the conversion chain. But one of the biggest uses might surprise you…

Up to half the energy of a train can be used just in stopping! Traditional brakes are a bit like the brakes on your bike, physically stopping the wheels with brake discs. The movement energy, which had first been converted from either electrical or chemical energy, gets wasted as heat in the brakes as the train slows down.

Axle mounted brake discs

Better Braking

Fortunately, there is a better way. Dynamic braking is a technology that allows the recovery of a vehicle’s braking energy as electricity. Instead of applying physical brakes, the electric motors act in reverse so that the force opposes the direction of rotation, slowing the train down and generating electric power. In some cases using dynamic braking can recover over a third of the total energy!

Normally, the extra electricity, known as regenerated energy, satisfies any immediate requirements on board the vehicle, powering comfort functions such as lighting, heating and cooling. But the destination of any surplus depends on the type of railway.

Regenerated Energy

In electrified networks, the extra electricity can be returned to the overhead line, or third rail, to be used by other trains. While this sounds great in theory, it requires a bit of thought in practice.

The national grid is a three phase alternating current (AC) supply system, carried in three wires. AC means that the supply voltage changes over time in the shape of a sine wave. Another supply type is direct current, or DC which provides a fixed voltage a bit like a battery. Electrical systems are connected to the electricity grid by one of three possible wire combinations, essentially meaning that there are three supplies. These supplies are really high voltage AC, and are usually converted to the appropriate type and voltage to power things. Typically, railways run on 25kV AC or between 600V-3000V DC.

As mentioned earlier, the demand from the rail network is huge, so if it were connected to just one of the three ‘supplies’ it would massively unbalance the grid! To overcome this, the railway is split into electrically isolated ‘sections’, connected to one of the three.

This means that if a train uses regenerative braking, the electricity it produces can only power trains within the same section. If there are no other trains in the section, the electricity is wasted - it heats up banks of resistors to prevent the line voltage from rising too high.

In non-electrified networks, diesel trains also use dynamic braking to prevent wear of the wheels and rails. But since there is nowhere for the excess electricity to go, it is diverted into resistors too.

Rail lines

How can we use it?

1. Timetable optimisation

There is one really easy way to ensure that we benefit from the regenerated energy in electrified networks: optimise the timetable so that when a train brakes there is another train to absorb the energy. This is particularly effective for metro systems, like London Underground, because there are lots of trains, all stopping and starting frequently. Under these conditions, even without optimisation, there is the possibility of energy exchange between vehicles, but the efficiency of doing so can be improved. Optimised timetables for metros in Beijing, Madrid and Xi’an provide estimated energy savings of between 5% and 15%!

2. Reversible substations

PylonsOne big limitation for regeneration on DC railways is that the electricity has to power a train close by, because there is nowhere else for it to go. The sections are powered by rectifier substations which only allow current to flow in one direction. But if those substations were to be ‘reversible’ the electricity could be returned to the national grid. In theory this means that all of the recovered energy could be used! Installation of this technology is expensive, but it has been done on both the Rotterdam and Bilbao metros, saving between 10% and 20%. The benefit of reversible substations is more significant in off peak conditions, when there is less chance of direct exchange between trains on the network. Substations on AC railways allow the regenerated energy to return to the grid without any special equipment.

3. Energy Storage

Both methods above assume that the energy has to be used immediately – but why not store it?

On board systems allow vehicles to store their braking energy until they need it later on. This has significant potential for diesel-electric vehicles especially, as they normally have to waste any regenerated electricity. However, there is a trade-off between storing all of the available energy and developing a system small and light enough to fit into a vehicle.

For electric railway networks, wayside energy storage is another option. Storage units can be positioned at fixed locations along the route, to supply energy to trains when they need it. Wayside systems aren’t really constrained by size and, in theory, can be as large as you want.

Batteries are the most familiar storage technology and there certainly are many kinds of batteries available: lead-acid, nickel metal hydride, lithium-ion and so the list goes on. But batteries are not the only option for rail. There are high speed mechanical flywheels, which store energy in a rotor in a vacuum, and special capacitors known as ‘ultra’ or ‘super’ capacitors. These work in a similar way to the capacitors you’d find in electric circuits, storing energy in an electrostatic field. All of these technologies have their pros and cons, which make storage design challenging. For each case the engineers have to decide which type, technology and size of unit is required.

Despite the difficulties in managing regenerated energy, operators are implementing all these methods with successful results, right now. They are also trying to save energy in a number of other ways, e.g., with more efficient equipment, eco-friendly driving, and better traffic management, to name just a few. The next time you want to make the ‘green’ choice, take the train. Railways really are driving us towards a more sustainable future.

London train rainbow


Subscribe Free

Related Content


Make a comment

Fill them with passengers. alancalverd, Fri, 2nd Oct 2015

Make them lighter would be a good start. Bored chemist, Fri, 2nd Oct 2015

How do they sync the phase of the regenerated supply coming off the train with the supply rail?
chris, Sat, 3rd Oct 2015

The presumption of "greenness" is false, surely.

To move people 200 miles, say from London to Manchester,  you need to destroy a strip of land 200 miles by at least 60  feet wide that cannot be crossed by wildlife, produces unacceptable noise pollution over at least 100 square miles, and involves the production and transport of thousands of tons of metal, concrete and gravel to make the infrastructure. You have to build new road and river bridges and tunnels and generally disrupt other traffic for about 5 years to build any railway.

The deadweight/payload ratio of a train is about 1 ton per passenger (Virgin Pendolino). A Boeing 737 travels at four times the speed of  a train, with one third of the deadweight per passenger, and requires no intermediate infrastructure - just a mile of runway at each end.

Since the passenger capacity of a Pendolino is about 500, and a 737, about 120, one aeroplane can replace one trainset and shift as many people per hour between London and Manchester without inconveniencing anyone on the ground in between. The service can start tomorrow, with no disruption of surface traffic.

The plane costs about 30 times as much as the train, but the infrastructure cost of a new railway completely dwarfs both considerations.

As for passenger convenience, the service interval is key. To shift the same number of people per day you would have to run 4 times as many flights as train journeys. So instead of, say, an hourly departure carrying 500 people for a 2-hour trip, you would have a flight every 15 minutes carrying 125 people for a 30 minute trip. Which is the more appealing, whether you are on business and have just  missed your train, or you have a squalling baby in arms and just want to get home asap? Or suppose you are on a business trip and the passenger next to you has a squalling baby?

One electrical fault, engine problem or dead sheep on the line will stop all trains for hours - and not only the Manchester shuttle. One faulty aeroplane does not prevent the 200 others from flying. If the runway at London City is congested, you can land at Stansted, Luton, Heathrow, Biggin Hill, Gatwick.... but you can't divert a train from Euston to Kings Cross. 

Airlines are not subsidised by the taxpayer.

So to make trains more efficient, put wings on them.

  alancalverd, Sat, 3rd Oct 2015

But Alan, airports aren't just about aeroplanes - you have to factor in the footprint of the airport itself, the huge ancillary parking and transport burden - all those people don't live at the airport, they need to move themselves there, often by road - and the convenience cost of regular transport movements (although where UK rail is concerned, I know I am on shaky ground here). chris, Sat, 3rd Oct 2015

But Chris, stations aren't just about trains - you have to factor in the footprint of the station itself (look at the marshalling yards outside all the London termini), the huge ancillary parking and transport burden - all those people don't live at the station, they need to move themselves there, often by road - and the convenience cost of regular transport movements.

The difference is that airports are usually outside town centres, so don't concentrate intercity traffic during rush hours when  the intracity traffic is at its peak. Next time you are in Manchester, Southampton, Leeds, or even London, compare the shambles around the railway station with the comparative calm of the airport (I'll admit the M25 around Heathrow is a mess, but Heathrow handled twice as many passengers as Euston last year). 

Given that we already have a few railways left after Beeching, the important question is what do we do next?If you really want to connect the "northern powerhouse" to London, don't waste money on a railway line that annoys everyone and takes forever to build: buy a plane and fly to Newcastle tomorrow. It's cheaper, too!  (£120 standard class to Kings Cross in 3 hrs, £75 to Stansted in 1:20)

Late edit: I just found some numbers....

So the proposal to spend £80,000,000,000 of your money on building a railway line to carry 3 trains per hour won't save any carbon emissions. The same money could have bought 1000 aeroplanes, 10 of which would have brought intercity travel in England into the 20th century. alancalverd, Sat, 3rd Oct 2015

Alan - can you help me with my query about syncing the phase of the regenerative output with the input line in systems designed to return electricity to the grid (as per what's in the article)?

C chris, Sat, 3rd Oct 2015

The problem is handled every day when generators are coupled to the grid. Not sure how they intend to do it with any particular train but one method (used with windmills) is to generate a DC voltage which is then "chopped" into sync with the nominal 50 Hz grid. If the synchronisation is correct, the generator will always be supplying power to the grid, so measuring the current flow thoughout the cycle will demonstrate any loss of synchronisation. 

However it may not be necessary to feed power back to the grid. Passenger trains have a significant demand for secondary power for air conditionng, lighting, etc., which must be maintained when the primary power fails (which it often does!) so it makes sense to use braking power to charge the backup batteries.

One advantage of DC power for trains is clearly that you don't need to sync the regenerative braking system to put power back into the grid. alancalverd, Sat, 3rd Oct 2015

The fixed-frequency AC grid is good for getting power from generators that run at a fixed speed, and transfer it to motors that run at an (almost) fixed speed.

This doesn't work so well for motors that need to run at a variable speed, like motors in lifts and trains.

So many electric train lines run off DC. Traditional DC electric motors can be wired to provide high torque at low speed, and the coils then reconnected by relays to provide more torque at higher speeds. A fairly simple rectifier circuit converts the AC mains to DC; but conversion of DC back to AC mains is so complex that it was not widely used, in the past.

Modern electronics have allowed the production of high-powered electronic inverter circuits that are able to convert DC to AC, or AC to AC of a different frequency and phase. Inverters with capacity of kW are required on the solar cells we see on the roof of so many houses. Inverters with capacities of many MW are needed for high-voltage DC transmission systems. Railway and wind power inverters are intermediate in capacity.

So trains can convert the DC to variable-frequency AC for the motors of an individual train, and also apply braking force, converting the kinetic energy of the train back into DC.

Large inverters at the substation can convert excess energy from the trains into fixed-frequency AC that matches the voltage frequency and phase of the AC grid. They can adjust the current phase to draw power from the public grid, or inject power back into the public grid. Agreement is needed with the local power utility to ensure that their design can deal with the fluctuating voltages and currents that this induces on their network. evan_au, Sun, 4th Oct 2015

I had a minor quibble with one part of the article:

In fact, for well-regulated DC, the rectifier needs to draw power from all 3 phases of the AC mains, so that when one AC phase is near zero (and can deliver no power), the other phases can deliver power into the train's DC power line.

But the overall conclusion was correct, in that the train power system is split into sections along the length of the track, and the best way to share excess power between segments is via the public AC grid.  evan_au, Sun, 4th Oct 2015

Yes, she says that the trains use 25kV AC supplies; so are you suggetsing that the AC is converted to DC on the train and then motors are DC? Or are the trains still directly consuming AC? What mechanism is therefore likely used to feed energy between the trains, like she says? Will they have high-power inverters on the trains? chris, Sun, 4th Oct 2015

HS2 and indeed most modern intercity systems use 25 kV AC. This has considerable advantages for transmission, load balancing and arc suppression (low voltage DC trains can really light up the city when there's snow on the rails!) but does make regen braking a lot more complicated. If it were my problem, I'd build gradient approaches to new stations, adopt run-and-glide timetables, and use any residual regen power to charge batteries on the train.  And of course regen braking won't actually stop the train: it only produces a braking effect when the train is moving! So the "green" answer, as with a car or a plane, is to modify your speed profile so you don't arrive with any spare energy. alancalverd, Sun, 4th Oct 2015

The use of AC or DC is a tradeoff which is up to the designer. Traditional electromechanical trains and trams often used DC.

Now that the price of high-power electronics has plummeted, it really doesn't matter whether the motors are AC or DC, or whether the supply voltage to the train is AC or DC - it is really quite easy to convert between AC & DC, and easy to convert from DC or fixed-frequency AC to variable-frequency AC.

With high-speed trains, what matters is that the voltage should be as high as practical, because the higher voltage transmits more power over longer distances with lower losses.

As Alan states, it is easier to build AC circuit-breakers.

It is also easier to convert AC at one voltage to AC at a different voltage, using transformers. If the train line overhead wires were powered by AC, it would be easier to power each section of each track from one phase of the grid. evan_au, Sun, 4th Oct 2015

hang on!, you can't say that and then quote a source for carbon emissions that says
"7 kilograms (15 lb) per passenger for conventional rail. Air travel uses 26 kilograms (57 lb) per passenger for the same journey. "
And that's before worrying about stratospheric pollution.

Also, re. " Next time you are in Manchester, Southampton, Leeds, or even London, compare the shambles around the railway station with the comparative calm of the airport (I'll admit the M25 around Heathrow is a mess, but Heathrow handled twice as many passengers as Euston last year).  "

Heathrow handles roughly twice as many passengers as Euston.
And it is 5km long and 2.5 km wide  ie 12.5 sq km
Euston is about 300 m by 500  ie 0.15 sq km

Did in to occur to you that part of the relative calm of the airport is simply because each passenger has eighty times as much space?

You have also forgotten that about 40% of air passengers at Heathrow got there on public transport- often by rail.
Heathrow is a pretty big railway station.

The UK's intercity 125 trains run on diesel.
They are, in effect, hybrids.
It's likely that any new rail links would be electrified and the plans for HS2 suggest 25KV AC- the same as most electric trains in the UK.
Bored chemist, Sun, 4th Oct 2015

Oops, sorry, only 40 items as much space I forgot about the passenger numbers. Bored chemist, Sun, 4th Oct 2015

You could design a passenger train as light as an aeroplane, but it would lose some of the safety of the traditional steel train chassis (and it might be more expensive than current train construction methods).

Planes are not exactly quiet, either...

Back to design tradeoffs; surely...

A train can save a fair amount of energy from frequent stops with regenerative braking. I have not yet seen a plane which regenerates av-gas when it descends towards the destination airport.

A plane is most economical when traveling at high altitude, due to reduced wind resistance. A train is pretty much limited to ground-level air pressure.

A train has an advantage carrying high-mass/low value goods. You don't normally weigh the passenger baggage on a train.

A plane has an advantage with low-mass/time-critical goods (at higher speeds).

The most dangerous location for a plane is when it is traveling within 10m of the ground/tarmac; the safest location for a train is when it is within 10m of the ground/rails.

So planes have an advantage for long routes where they spend a small fraction of it near the ground; and trains have an advantage for very short routes between points on the ground.

Occasionally, proposals surface for vacuum tube or pressure tube trains, which might shuffle some of the above priorities, but at greatly increased capital cost and safety risks... evan_au, Sun, 4th Oct 2015

So just to close the loop on my question (sorry, but we keep skirting the issue) in case someone else is reading this:

You're theory would be that the train will probably have inverters on it that match up the "regenerated" electricity with the supply phase and then inject the current in sync? Or it's DC in the first place and then relatively easy to sync it up.

C chris, Sun, 4th Oct 2015

Chris: yes. If it's an AC supply you need a synchronous inverter, with DC supply things are a bit easier, but either way you are relying on the elasticity of the grid to absorb the braking power.

BC: Heathrow is a very oldfashioned airport, built in the days of marginal takeoff and crosswind performance when you needed at least three 60-degree runways and swampland was cheap. London City is an example of what can be achieved: transatlantic and commuter flights from less than a quarter of a square mile - it's not much bigger than Kings Cross.

As for carbon emissions, you pays your money and takes your pick. The faster you go, by whatever means, the more juice you burn. But it's important to consider the lifetime emission of the whole system. Raw steelmaking produces about 2 tonnes of CO2 per tonne of steel, and you can probably add another 0.5 for transporting and reforming it into a product in place. Two railway tracks weigh 1660 tonnes per mile, so just making the rails for a 200 mile track from London to Manchester will emit  830,000,000 kg of CO2, enough to transport 62,000,000 passengers by air over the same distance in less than half the time. alancalverd, Sun, 4th Oct 2015

Apples<> oranges
You seem to have overlooked the considerable amount of metal used to make aeroplanes from. Plenty of our rolling stock is old- decades old. I know three are some old planes out there but they are to some extent like grandfather's axe. It's had three new handles and two new heads, but it's still going strong after all these years. So I rather suspect that making planes uses a lot of energy too, especially if you wan twelve billion passenger miles.
A full cost benefit analysis is beyond the realms of a discussion site like this but, as you say, going fast needs a lot more fuel.

That's rather at odds with the claim that "The presumption of "greenness" is false, surely."

Bored chemist, Sun, 4th Oct 2015

OK, the easy way - and probably the only effective way - to make trains greener is to make them slower.

You may be able to improve blunt-face drag by fitting suburban commuter trains with a false nose, but that introduces complications when you want to couple them together or negotiate sharp bends, and it doesn't affect the main fuselage drag. High speed trains are probably as drag-efficient as you can make them without compromising safety and ride quality. If you want them to go fast they either need a heavy undercarriage or steeply banked curves, but you can't run freight traffic on a steep bank, so we are back to destroying the countryside and making another load of steel just for passenger trains. That philosophy worked for the TGV and various Asian bullet trains, but nobody ever claimed it was a green alternative.

So the question is how much energy can be conserved by regenerative braking? Once again, it's all down to speed.

If you just switch off the engines of a 737, it will glide 85 miles from 30,000 ft and land at a comfortable 120 knots. The additional journey time for a 200 mile trip works out at about 5 - 10 minutes compared with maintaining cruise power up to final approach.

So what happens if you wind your train up to 200 mph, switch off the power, and coast down to 50 mph? That's a lot more efficient than trying to pump energy back into the grid in order to decelerate quickly. So ideally, you would only use regen braking as an adjunct to friction braking, in an emergency (remember it doesn't work well at low speeds). But if your friction brakes are adequate, you can save weight by eliminating the regen system, and enhance friction braking by opening slots on your aerodynamic fairings to slow down.  Regenerative braking is a Good Thing for short journey legs, so great for road traffic and even underground trains, but not the obvious solution for intercity trains.

Apropos the CO2 cost of building an aeroplane: Assume it is all made of aluminium (the most energy-intensive component). That requires about 12 tonnes of CO2 per tonne of finished product. So making one 80 tonne 737 generates 960,000 kg of CO2 - enough to make about 0.25 miles of railway track. It will cover about 10,000,000 miles in a conservative service life of 20,000 hours (modern jet engines last about 10,000 hours and you'd probably replace the soft furnishings at the same time, but a modern metal hull is pretty much indestructible). A 400-tonne train requires about 800,000 kg of CO2 to manufacture and will probably last 40 years and cover half that distance.

So the train is marginally greener as long as it can run on existing tracks, but as soon as you build or replace a couple of miles of track, it would have been better to use a plane. alancalverd, Sun, 4th Oct 2015

Excellent news. How would you make trains more energy efficient? alancalverd, Wed, 7th Oct 2015

Wow! Talk about bearing a grudge - you deserve a medal!

Maglev certainly addresses the problem of maintaining stability without increasing frictional losses, but do you have any energy data on the systems? I had some fun with early Laithwaite linear motors but only using resistive magnets to shift very small loads - I don't think the word "efficiency" ever passed our lips. Supercons make great headlines but my experience with high temperature supercons in MRI is that the cost (i.e. energy consumption)  of continuous refrigeration when the device is idle, exceeds the cost of powering up a resistive magnet of the same field strength for a 50% duty cycle, and in practice it's just one more bloody thing to go wrong (as an engineer you will appreciate this).

Pity about the Advanced Passenger Train, but the whole business of high speed rail travel over short distances surely needs rethinking. If you travel between London and Birmingham at 100 mph it will take about an hour. At 125 mph the same train will save 12 minutes but will need 1.5 times as much power and expend 1.25 times the energy to overcome the aerodynamic drag.  Increasing the speed to 150 mph requires another 44% power increase to save a further 8  minutes....a seriously big engine to get a seriously small return! alancalverd, Thu, 8th Oct 2015

Maglev trains are actually more efficient than normal trains. The energy needed to lift them is smaller than their air conditioning takes, and they get to the destination slightly faster, which also cuts the energy used. They also have lower wear and tear on the track and are less prone to 'wrong kind of snow' issues.

The downside is that their tracks are about a couple of times the cost of conventional trains, so they're expensive to install. wolfekeeper, Wed, 21st Oct 2015

I recently spent four an a half hours driving from Maidstone to Heathrow to collect my Son flying in from Frankfurt, air travel for a few hundred miles is one of the slowest and miserable forms of travel possible.
I told him to use the train next time. syhprum, Wed, 21st Oct 2015

Apples and oranges! It takes about 15 minutes to fly from Maidstone to Heathrow, and between 7 and 12 hours to travel from Maidstone to Frankfurt by train or car. You could fly in a small plane from Rochester to Frankfurt in about 2 hours, or in a jet from Manston in less than 1 hour.  The problem isn't air travel, it's the pathetic surface travel in and around London, the fact that Manston was sold for £1 to build houses under the government's "brown field " scheme, and the absurd tax on fuel for piston-engined aircraft.

Next week I have to travel between Cambridge and Llanelli. 4.5 or 7.5 hours each way by train (plus 40 minutes from any Cambridge "park & ride" to the station in the rush hour), 8 hours by bus, 5 hours driving, or 90 minutes in a 2-seat basic training aircraft. No contest. alancalverd, Wed, 21st Oct 2015

Personally, I don't believe that aeroplanes can ever work. If you did somehow take off you would have to land and aeroplanes would take miles to stop.

That would mean that aeroplanes would often have to be quite a long way away from centres of cities, which would make them extremely inconvenient.

Clearly, man will never fly.

Oh wait, they do fly, but that's why they're not used to go down to the shops.

My bad! wolfekeeper, Wed, 21st Oct 2015

See the whole discussion | Make a comment

Not working please enable javascript
Powered by UKfast
Genetics Society