(468e) Modeling and Control of a Hybrid Renewable Energy System

The utilization of renewable energy sources is becoming increasingly attractive and widely used as an alternative to oil-produced energy. Due to the widespread availability of renewable energy (RE) sources, power can be produced in a small scale and remote locations, as long as a sustainable integrated system for power production, storage and distribution exists. Energy storage is a key challenge, especially when the energy system is not connected to the grid, and is usually solved by incorporating an energy storage unit. For long-term storage, electrical energy can be converted into hydrogen, for later use in fuel cells to produce electricity.

This study deals with the development of a supervisory, optimization and power management tool for a 10kW hybrid energy system built in Lambton College, Sarnia, ON.  The main rationale for such a tool lies in the fact that renewable energy sources are intermittent in nature and their availability may not coincide with the load demand or with the availability of energy carriers. In order to accomplish the task of developing the supervisory software, the main step is to develop a reliable plant model, which is then used for optimization and control purposes.

The wind, PV, and hydrogen production/storage/consumption components of the hybrid energy system are modeled individually. Process streams and electrical connections are applied for integration of the above-mentioned components. The RE sources and fuel cell are integrated either through an AC or DC bus. The hybrid energy system is modeled as stand-alone and grid-connected as well. When the hybrid system is without grid connection, the power demand (load) is the main criterion for power management. However, in the grid-connected system, power can be transferred from the grid to the hybrid system when power generated from RE sources does not meet the load demand. The integrated model was structured in a dynamic mode. In addition to main components, hydrogen storage tank, gas compressor and gas pressure regulator models are also developed.

The hybrid system is designed and modeled in a modular approach for all components, which allows proper sizing of equipment to achieve maximum efficiency. The control architecture consists of the low level local controllers for each component and a supervisory power controller, which manages the power flow. Excess RE power initiates the hydrogen production units and whenever the RE system cannot completely meet load demand, the FC system provides power to meet the remaining load. The supervisory controller deals with a complex control problem of handling binary and continuous variables, as well as constraints. The control strategy consists of tracking the power demand, producing sufficient amount of hydrogen, and switching on and off modular components. It is important to highlight the consideration of constraints such as minimum fuel cell, electrolyzer power threshold and time limitation between fuel cell and electrolyzer startups and shutdowns. A model predictive control strategy is developed to solve the optimal energy management problem.