This thesis is concerned with the topic of increasing energy efficiency in DC railway systems, to reduce CO2 emissions associated with this form of transportation by minimising the energy consumption and energy loss of the system. In this work a method is proposed and implemented to model a DC railway that enables the effects of incorporating an energy storage system (ESS) to improve energy efficiency to be investigated. To date, no systematic investigation has considered the parameters by which an ESS can be controlled, the effects this has on the power flows of the electrical system and hence be able to provide an accurate energy analysis.
Electric trains are capable of regenerating energy that can be utilised by other trains operating on the same conductor rails, reducing the traction energy demand by up to 50%. However, poor synchronisation between accelerating and decelerating trains causes an excess of regenerative energy to be dissipated in braking resistors in the form of heat. ESSs have been proposed to assist in the energy efficiency improvement of electric railways by exploiting the captured excess braking energy to reduce the traction energy demand.
A simulation model for an electric railway system was developed in MATLAB software to investigate the application of a generalised ESS to the railway model. A comprehensive sensitivity analysis was carried out to explore the optimal voltage thresholds of an ESS controller that could maximise the energy savings in the electric railway system. The results show that optimising for the best energy savings will lead to a significant imbalance of the ESS import and export energy, meaning that the ESS will always trend to reach its state of charge (SOC) boundaries. As an ESS cannot have infinite capacity, and practical restrictions mean that its SOC cannot be separately managed, under normal operation the ESS would always become unavailable negating any positive effect it would have had.
A multi-objective algorithm was used to select the optimal voltage thresholds for an ESS controller, with fixed current control, and simulated under deterministic behaviour of the railway system operation. This was then extended to investigate two adaptive control methods to dynamically change the ESS current in response to variable railway operations. The first adaptive control method used state machine control while the second method employed fuzzy logic control to adapt the ESS current to avoid ESS unavailability without compromising on the energy savings. The proposed control methods were implemented experimentally for validation.
Three case studies representing different railway operations were applied to the electric railway model and then simulated for two days to test the impact of the proposed control methods on the energy savings. The results indicate that the ESS controller applying the optimal voltage thresholds and fixed current provided energy savings of 42.34%, 41.02%, and 30.18% when simulating the first, second, and third case study, respectively. The first adaptive control method implementing state machine achieved energy savings of 38.96%, 41.55%, and 38.75%, while the second adaptive control method implementing fuzzy logic yielded 41.43%, 43.08%, and 39.21%, respectively. The proposed control methods were validated experimentally to prove their capability in real-world applications. The experimental results of the proposed controllers applied to the three case studies were compared against the simulation results, showing a percentage error ranging from 1.8 to 3.6%.
The findings reported in this thesis suggest that ESSs can effectively import and exploit the regenerative energy to improve the energy efficiency of electric railways. Moreover, the benefits of ESSs could be maximised while avoiding their unavailability through adaptive control and careful selection of the ESS operating parameters.
