A New Generation of Modular Multilevel Converters Integrating Energy Storage Devices for Dual-Voltage Railways

Historically, supply voltages of traction systems were developed independently in different parts of the UK and in different countries. This led to the proliferation of several types of railway electrification systems, either DC or AC. Nowadays the permissible range of voltages of traction systems are set out in the standard BS EN 50163. Due to the high capital costs involved, it is unlikely that a single standard railway electrification system will be adopted in the near future. The large range of voltages and the AC and DC supply require that services operated across more than one electrification system are operated by rolling stock equipped by dual-voltage traction drives. Since the formation of the European Union and the subsequent increase in the amount of cross border traffic, new needs for multi-voltage trains have arisen.
Traditional traction converters have a DC bus and provide three-phase power for the traction motors. The DC bus voltage is usually the same of the DC power line and, hence, traction converters are directly connected to the DC supply. For AC supply, the direct connection is not possible and an additional AC/DC four quadrant converter is needed. Moreover, a fault in one of the switches affects the functionality of the whole traction converter. When energy storage devices are used, the voltage level of the DC bus requires necessarily the connection in series of the storage cells and, hence, the presence of balancing circuits and balancing control.
Traction converters for dual-voltage trains would benefit significantly by the introduction of new topologies with "universal" input voltages, because the AC/DC four quadrant converter could be removed. Modular multilevel converters present this characteristic and can be designed for a wide range of input and output voltages. Their topology is intrinsically fault-tolerant and the replacement cost of one module is significantly lower than that of the whole converter. Each module has an accessible DC link that gives the possibility of the integration of the storage elements without their direct series connection. Thus there is now a critical opportunity to develop a dual-voltage traction system with modular multilevel converters and integrated storage devices.
A major technical hurdle for the widespread application of dual-voltage trains with integrated storage devices is that the controllers of the converter have to be properly modelled and characterised if to be used for traction applications. The research group at the University of Birmingham will take the timely step of investigating for the first time how a modular multilevel converter with integrated storage devices can improve the performance of dual-voltage trains to benefit the connections between different electrification systems in the UK and the rest of Europe. A novel controller, based on extensive modelling of the converter, will be created to obtain optimal torque performance for the full range of the output frequency. A kW-range demonstrator with a modular multilevel converter and supercapacitors will be designed and constructed. The experimental results will give us confidence in applying the design methodology for future designs of large-scale dual-voltage converters. The research outcomes will deliver specific recommendations to train manufacturers for complementing traditional traction drives with flexible and modular solutions.

This project tackles the challenges of storage device integration in dual-voltage traction systems and provides strategic direction for future railways, delivering tools and technologies for the transition to low-carbon trains. The potential beneficiaries are manufacturers of rolling stock and traction systems, rolling stock owners (leasing companies in the UK), train operating companies, electricity suppliers and academia. The immediate outcome of this project will be to contribute to the achievement of the 4C targets (Cost, Carbon, Capacity, Customers), which have been identified as a priority by the DfT through the Rail Innovation Fund of the Enabling Innovation Team.
In terms of Cost, the research will enable companies to improve the availability of trains and reduce delays in case of a fault of the traction system. This will bring down the penalties applied for causing disruption in the timetable and increase customer satisfaction. The new modular converters will also reduce the maintenance costs of rolling stock traction systems, because faulty modules can be easily replaced at a cost much smaller than that of the whole traction converter. This will bring down the risk and therefore the cost of rolling stock over its lifetime. Electricity suppliers will also benefit from this research, because the better quality of current drawn by the trains will require smaller filters in substations.
As far as Carbon is concerned, the research will provide a platform to test the effectiveness of energy storage devices, with potential future expansion to new developments in low carbon technology such as hydrogen fuel cells. In the longer term, flexible topologies of multilevel converters with integrated storage devices and low-carbon sources will have a significant impact on the management of partially electrified lines, enabling the conversion of diesel rolling stock to new environmentally friendly trains.
In terms of Capacity, the capability of having a dual-voltage supply for the traction system will provide more homogeneity among vehicles of different types operating on the same line, with consequent better exploitation of the line capacity; in fact, vehicle heterogeneity accounts for significant capacity reductions on busy lines. The storage devices and the reactive power and power factor control of modular multilevel converters will reduce the power peaks and the distortion of the currents drawn by the trains for both DC and AC supply, helping towards improvement in the efficiency of the energy supply.
In terms of the Customer as a Train Operating Company, the vehicle leasing cost should be lower, as rolling stock will remain operationally viable for its full life, the system will be fault tolerant, and the replacement of modules will be quick and easy. As far as train owners or leasing companies are concerned, their rolling stock will be future proofed, and they will have the ability to adapt and match the vehicle's tractive capabilities to changing market needs, or to operate efficiently on different routes or service patterns.