Stochastic scheduling of self generated energy: Case for organized industrial zone

As a result of COP21 meeting in Paris in November 2015, 138 countries acknowledged that fossil energy has a major role in the global warming problem due to greenhouse gas emissions. The global use of renewable energy sources is increasing as one of the solutions for mitigation of warming effects. Hence, renewable energy resources become widespread in cogeneration and trigeneration systems to replace natural gas. Energy consumption in manufacturing industry has been increasing because of the technological improvements. The industry can reduce costs by generating energy to respond some of the self-demand. Cooling is another important need in industry besides power and heating. Despite all these advantages of trigeneration, it has not become frequent in heavy energy using industries. A hybrid trigeneration system that utilizes renewable energy resources such as sun and wind, and also natural gas and grid -if required- is studied in this thesis. The aim of this study is to propose a scheduling model for hourly energy generation and consumption system in an industrial center. To constitute the mentioned scheduling model, stochastic approach is employed due to the uncertain parameters of solar irradiation, wind speed, energy demands, and power purchasing price. Sampling Average Approximation (SAA) method, which is one of the scenario-based stochastic optimization methods, is chosen to handle with these uncertainties. High variance of the optimality gap in SAA leads to a significant problem. The variance reduction techniques have been examined in order to solve it. Considering the uncertain parameters and the structure of the problem, auto-regressive integrated moving average (ARIMA) method, which is a stochastic process with discrete time and continuous state space, is selected among the variance reduction techniques. Then, a hybridized model is designed as "ARIMA-aided SAA". All the steps of SAA method have been carried out again after the improvement with ARIMA. The proposed stochastic model is applied for a hybrid trigeneration system in Gebze Organized Industrial Zone (Gebze Organize Sanayi Bölgesi - GOSB). The classical SAA method and the hybrid method of SAA and ARIMA are implemented for the mentioned zone with different sample sizes. The results are evaluated in terms of the optimum objective function value, the optimality gap, and the variance of the optimality gap. The presented industrial trigeneration system is characterized by multi-source energy generation to reduce the climate change effects, will have impacts on reduction of natural gas importation. The industrial center will reduce the uncertainty problems by generating the energy to respond its own demand. Furthermore, a new opportunity is provided to the center by the help of hourly schedule for the energy generation and consumption.