Search

Incorporation of renewable sources and energy storage can contribute to reduce fuel consumption in diesel generation-based remote off-grid power systems. This thesis proposes methodologies for improved management of available generating and storage resources to reduce costs and emissions and investigates the coordination of energy management with power management functions essential for stable operation of the isolated power systems. Firstly, a test system is developed considering a case of retrofitting an existing diesel power system with photovoltaic (PV) generation and battery energy storage. A sizing study conducted using HOMER software with weather and load data for Northern Canada resulted in a PV-Diesel-Battery topology with high PV penetration levels. Next, the energy management functions are developed incrementally in three steps. First, a computationally efficient energy management system (EMS) is implemented to optimize the system operation while incorporating multiple operational requirements essential to remote power systems. Then, a demand response (DR) model that requires minimal bi-directional interactions and therefore implementable without sophisticated communication infrastructure is developed and integrated with the EMS. Thirdly, a computationally efficient two-stage model predictive control process is developed to compensate for forecast uncertainties. Finally, the power management functions necessary to achieve a logical real-time operation are implemented and coordinated with the energy management functions in a hierarchical architecture. Also, an operation evaluation framework is suggested to assess the viability of the optimum operation routines under dynamic conditions. After adding PV and energy storage to an existing diesel-only system and optimizing the operation with the DR integrated EMS, over 60% cost and emission reductions are achieved for a representative summer day compared to the diesel-only operation. The cost and emission reduc