What is Regenerative Braking? Energy Recovery in Trains

A Tokyo metro train approaching Shinjuku station at 80 km/h carries approximately 150 megajoules of kinetic energy in its 300-tonne mass — roughly equivalent to the energy content of 4 litres of diesel fuel. In the 30 seconds it takes to decelerate to a stop at the platform, that energy must go somewhere. In a friction-braked vehicle, it becomes heat — the brake discs and pads convert kinetic energy to thermal energy that radiates into the tunnel environment, contributing to the notorious underground heat problem of metro systems. In a regenerative-braked vehicle, the same energy becomes electricity — fed back into the overhead supply, where it is immediately available for the train accelerating out of the previous station.

Regenerative braking is one of the oldest energy recovery technologies in electric transport — first used on electric tramways in the 1890s — but it has become a central pillar of sustainable railway operations as the industry seeks to reduce both energy costs and carbon emissions. Understanding how regeneration works, where the energy goes, and why it sometimes cannot be used is fundamental to understanding the energy economics of electric rail.

What Is Regenerative Braking?

Regenerative braking exploits the reversibility of an electric motor. When powered electrically, a motor converts electrical energy into mechanical rotation (traction). When driven mechanically — as happens when a moving train’s wheels turn the motor shaft during braking — the same machine operates as a generator, converting mechanical energy into electrical energy.

The train’s power electronics control this transition. During normal traction, the inverters supply variable-frequency AC to the motors. During regenerative braking, the same inverters reverse the energy flow — the motors generate AC, which the inverters convert to the traction supply voltage (DC or AC depending on the system) and feed back into the contact wire or third rail.

The braking force produced by regeneration is a result of the electrical resistance the generator presents to the wheel rotation — essentially, the motor “pushes back” against the rotation, slowing the wheels. The harder the motor works as a generator (the more current it produces), the greater the braking force.

The Four Braking Modes: From Recovery to Friction

Where Does the Regenerated Energy Go?

Regenerated energy has three possible destinations, listed in order of preference from most to least efficient:

1. Absorbed by an Adjacent Accelerating Train

This is the ideal scenario: the regenerating train feeds energy into the contact wire while another train on the same electrical section is simultaneously accelerating and consuming energy. The energy flows electrically from one train to the other without any conversion losses beyond transmission resistance. On a dense metro system with trains every 90 seconds, this scenario is common during peak hours — at almost any moment, a braking train and an accelerating train are sharing the same electrical feed section.

The efficiency of train-to-train energy transfer depends on the synchronisation of braking and acceleration events and the electrical resistance between the two trains. On a DC system with closely spaced substations, the resistance is low and transfer efficiency is high — typically 80–90% of the generated energy reaches the consuming train.

2. Stored in On-Board or Wayside Energy Storage

When no adjacent train is available to absorb regenerated energy, storage systems can capture it for later use:

• On-board batteries: Lithium-ion battery packs on modern trams and BEMUs store regenerated energy and release it during subsequent acceleration. On-board storage is particularly valuable on non-electrified sections, where the stored energy enables catenary-free operation.
• On-board supercapacitors: Supercapacitors have higher power density than batteries — they charge and discharge faster — making them better suited to the rapid energy pulses of urban rail braking and acceleration. Several Siemens Desiro and Alstom Citadis tram variants use on-board supercapacitors.
• Wayside energy storage (WESS): Battery or supercapacitor banks installed at substations capture regenerated energy when no train is available to absorb it, and release it during the next high-demand period. Madrid Metro, Barcelona Metro, and Paris RER have deployed WESS with documented energy savings of 20–35% at equipped stations.

3. Fed Back to the Grid via Reversible Substations

On AC 25 kV systems, modern four-quadrant converter technology allows regenerated energy to flow naturally back through the traction transformer to the grid. On DC systems (750 V or 1,500 V), conventional diode rectifiers are one-directional — they cannot pass energy back to the grid. Reversible substations with active front-end inverters are required to enable grid feedback on DC systems.

Grid feedback is less efficient than direct train-to-train transfer (additional conversion losses) but more efficient than storage (avoids storage round-trip losses). It also provides a valuable ancillary grid service — if the railway’s regenerated energy is exported to the grid during peak demand periods, the infrastructure manager may receive revenue from the grid operator for providing demand response.

Regenerative vs Rheostatic Braking: Full Comparison

Factors Affecting Regeneration Efficiency

The actual energy recovered by regenerative braking varies significantly between systems and operating contexts. Several factors determine how much of the theoretical regeneration potential is actually realised:

Traffic density: The probability that regenerated energy can be immediately absorbed by an adjacent accelerating train increases with traffic density. On a metro running at 90-second headways, the probability of a simultaneous braking/acceleration pair is very high. On a mainline with 15-minute headways, it is low — most regenerated energy must be stored or wasted.

Service pattern: Stop-start urban services generate more braking events per kilometre than high-speed intercity services, creating more regeneration opportunities. A metro train brakes and accelerates approximately every 2 km; a high-speed train may travel 200 km between significant braking events.

Gradient profile: Descending gradients create sustained regeneration opportunities — a train coasting downhill at constant speed is effectively generating electricity. Ascending gradients eliminate regeneration entirely (the train is accelerating, consuming power). Routes designed with “humped” profiles — with stations at summits and running sections in valleys — maximise regeneration by having trains accelerate downhill from stations.

Supply system receptivity: On DC systems, if no train is available to absorb regenerated energy and no storage is present, the DC busbar voltage rises. When it exceeds the maximum safe level, the on-board control system must switch from regenerative to rheostatic braking to protect the supply equipment. This “supply overvoltage” condition is common during off-peak hours when train density is low.

Real-World Energy Savings: Documented Case Studies

Timetable Optimisation: Making Regeneration More Effective

One of the most cost-effective approaches to improving regeneration utilisation requires no hardware at all — just better timetabling. By adjusting departure times so that trains on adjacent sections are braking and accelerating simultaneously, operators can maximise the probability of train-to-train energy transfer.

Hong Kong’s MTR Corporation has implemented energy-optimised timetabling that intentionally staggers departures to create braking/acceleration pairs on the same electrical section. The approach requires sophisticated simulation software to identify the optimal departure timing, but the hardware cost is zero — only the software and operational discipline to maintain the timetable. MTR reports consistent energy savings of approximately 25% from regeneration, compared to 15% before timetable optimisation.

The Tunnel Heat Problem: Why Regeneration Matters Beyond Energy Cost

In deep-bore metro tunnels, heat management is an engineering challenge of growing severity. Trains generate heat through traction motor losses, air conditioning rejected heat, passenger body heat, and — on non-regenerative systems — braking. In a sealed tunnel environment, this heat accumulates over time, raising tunnel temperatures to levels that affect both passenger comfort and equipment reliability.

London Underground’s deep tube lines are a well-documented example: despite relatively low ambient temperatures at surface level, deep tube tunnel temperatures have risen by approximately 1°C per decade since the 1950s, as the clay surrounding the tunnels gradually absorbed heat from train operations and reached thermal saturation. Modern rolling stock with full regenerative braking contributes significantly less heat to the tunnel environment than older stock with rheostatic braking — one of the operational benefits of fleet replacement on deep metro systems, beyond direct energy cost reduction.

Editor’s Analysis

Frequently Asked Questions

Q: Does regenerative braking fully replace friction braking on trains?

No — regenerative braking is always used in conjunction with friction braking, not instead of it. Regenerative braking is most effective at moderate to high speeds, where the motors can generate significant braking force. At very low speeds (below approximately 5–10 km/h), the regenerative braking force drops to negligible levels and friction brakes must take over for the final stopping movement. Additionally, friction brakes are required for emergency stops where maximum deceleration is needed regardless of electrical conditions, and as a backup if the electrical system fails. Modern trains use “blended braking” — automatically combining regenerative and friction braking to optimise energy recovery while meeting the required deceleration profile.

Q: Why can’t all the regenerated energy always be fed back to the grid?

On AC 25 kV systems, modern four-quadrant converters allow regenerated energy to flow naturally back to the grid through the traction transformer — there is no fundamental barrier. On DC systems (750 V third rail or 1,500 V overhead), the rectifiers that convert grid AC to traction DC are typically one-directional — diode rectifiers that cannot pass energy in reverse. When regenerated energy enters the DC busbar with no adjacent train to absorb it, the busbar voltage rises. If it exceeds the safe limit, the regenerating train must switch to rheostatic braking. Reversible DC substations with active front-end inverters solve this problem but require capital investment that not all operators have made.

Q: How much money does regenerative braking save in practice?

The financial saving depends on the system’s energy price, traffic intensity, and the efficiency of energy recovery. For a medium-sized metro system consuming 200 GWh per year at €80/MWh, a 25% regeneration saving represents €4 million per year in avoided energy costs. For a national railway network consuming 2,000 GWh per year, the same efficiency gain is worth €40 million annually. Over a 30-year infrastructure lifecycle, these savings are substantial relative to the capital cost of enabling technologies (reversible substations: €2–5 million each; wayside storage: €1–3 million per installation).

Q: Do diesel trains use regenerative braking?

Conventional diesel trains do not use regenerative braking in the same sense as electric trains, because there is no electrical supply to feed energy back into. However, diesel-electric locomotives and DMUs with electric transmission can use rheostatic braking — the diesel engine drives a generator that powers electric traction motors, and during braking those motors operate as generators, dissipating the energy in roof-mounted resistors. Modern bi-mode trains and hydrogen trains with battery buffers can recover some braking energy into their on-board storage, but the total energy recovered is limited by the storage capacity rather than the braking force available.

Q: Is regenerative braking better on downhill sections?

Yes — significantly. On a descending gradient, the train’s potential energy is continuously converted to kinetic energy, which can be captured as electrical energy through regenerative braking. A train descending a 20‰ gradient at constant speed is effectively generating electricity the entire time — the motors provide continuous regenerative braking to maintain speed rather than accelerating. Routes with significant vertical relief (mountain railways, routes with valley stations) can achieve much higher regeneration rates than flat urban metro routes. The Rhaetian Railway in Switzerland, which descends hundreds of metres between mountain stations, generates significant surplus electricity from regenerative braking that is exported to the grid — the railway is effectively a hydro-power generation system as well as a transport system.