Regenerative braking overvoltage on a line with no receptive load

During regenerative braking the traction motors act as generators and the converter tries to push power back into the 1500 V DC line. With non-reversible rectifiers and no other train drawing power, there is almost no receptive load. The only sinks are line capacitance, train auxiliaries, and losses, so the line voltage rises quickly. When it reaches 1800 V the converter trips regeneration and blends to friction braking.

The repeated alarms mean the train is spending more time on friction brakes than designed. Pad wear of 3.0 mm per 1000 km versus 1.0 mm is consistent with that. The “short loss of brake blending” matters because a badly managed transition can produce a jerk or a temporary reduction in total braking effort; the friction brake must pick up smoothly and with enough margin.

The auxiliary inverter trip at 1850 V is the more serious warning. It shows that the overvoltage is not confined to the traction converter: auxiliary equipment and its protection are also being exposed. Raising the regen threshold to 1900 V would allow even higher line voltage and could reduce the margin for auxiliaries, insulation, and other trains' converters. It treats the alarm, not the cause.

Raising the threshold to 1900 V. Low cost and reduces nuisance trips, but does not create receptivity. It raises voltage stress on auxiliaries, insulation, and other rolling stock. The 1850 V auxiliary trip shows the margin is already thin. It may be acceptable only as a small, validated adjustment after insulation and auxiliary checks, not as the main fix.

Braking resistors. Proven and independent of the infrastructure. They dissipate the 2.5 MW as heat, so tunnel or depot ventilation and fire risk must be assessed. They add weight and cost to the train but guarantee braking energy disposal. Good fallback where receptivity cannot be assured.

Wayside storage or reversible substations. They recover energy and reduce line voltage rise, but are capital-intensive and need space, protection changes, and a business case based on energy savings and duty cycle. They are most attractive on sections with frequent braking and few receptive trains.

Timetable changes. Keeping another train in the section can provide receptivity, but it is fragile and may conflict with the service pattern. It is an operational mitigation, not an engineering solution.

The best choice depends on the whole route. A single 7 km section with low night traffic may not justify reversible substations; a braking resistor or controlled blending is more robust.

Do not accept raising the threshold to 1900 V as the primary remedy. The evidence points to lack of receptivity, and the auxiliary inverter trip shows the present upper voltage already stresses equipment.

Recommended response:

• Keep the regenerative braking threshold at a value that protects auxiliaries and insulation, and improve the blending strategy so that friction braking is applied progressively before the converter trips. The blend should be bumpless and should not depend on a sudden regen failure.
• Assess a braking resistor or a wayside storage/reversible substation for the section, using the actual braking energy and number of brake applications. If the business case is weak, the resistor is the more predictable solution.
• If a threshold increase is still proposed, require evidence that the traction converter, auxiliary inverter, protection settings, and insulation coordination all have margin above the new value under worst-case no-load conditions. A small step, for example 1850 V, is not justified merely to reduce alarms.

Verification:

• Record line voltage, regen power, and blending behaviour on the section at night with one train.
• Confirm no auxiliary overvoltage trips over a defined test period.
• Measure friction brake pad wear after the change to confirm the regen share has actually improved.
• Check that protection settings and other trains' converters are not exposed to the new maximum voltage.