SIMULATED PERFORMANCE OF A HYBRID ENERGY SYSTEM FOR A REMOTE AREA

Abstract

In India, as a developing country, the concept of providing sustainable power supply is required in order to provide electricity to remotely located villages which have no access to grid supply due to difficult terrain. Conventionally, remote consumers are supplied by diesel generator based system which operation is constraint by depleting nature of fossil fuel reserves, and negative social and environmental impacts. Recently, the concept of providing electricity to remote inaccessible locations through locally available renewable energy based power systems is being increasingly recognized and gaining interest globally. Among the available renewable energy sources, solar and wind are considered as the major energy sources in satisfying future electricity demand. Such renewable energy sources are environmentally friendly and available free of cost. However, the stochastic behavior inherent to renewable energy sources such as solar and wind pose serious challenges in designing power systems based on renewable energy sources. One excellent solution for nullifying the impacts of intermittency coming out of experience is the hybrid integration of two or more different energy sources so that the energy generated from different sources are complemented each other. But, the combination of different energy sources to form a hybrid system not only increases the complexity of the system , but also complicates the system optimum design process on account of unpredictable renewable energy sources, time changing load demand, system components non-linear characteristics and, the interdependent behavior of the system optimum configuration and optimum control strategy. These complexities introduce more difficulties in designing and analysis of hybrid systems. Hence, a comprehensive study related with different aspects of hybrid energy system is essential before any plan is executed. The current thesis focuses on the development of a new approach for sizing of a hybrid energy system consisting of solar, wind and battery storage with the objective of minimizing total system cost and supplying the required demand reliably. To carry out the study, a case study area located in a remote hilly region of Uttarakhand, India is considered and required hourly renewable energy inputs data are synthesized using the available monthly average solar radiation and monthly average wind speed. The load profile is generated using the seasonal peak energy requirement. Designing the techno-economic aspects of a hybrid energy system for remote area applications has showed a remarkable increase recently and consequently determining the system design which is optimal in terms of operation and component size is much crucial in order to have reliable and economical power supply. Owing to the fluctuating nature of iv | P a g e renewable energy sources, correct sizing of hybrid energy system require accurate modeling of its components so that their power output can be estimated exactly. Reported literatures suggest various techniques for modeling of each hybrid energy system component. However, they either adopted deterministic or probabilistic methods. In the current study, photovoltaic component of hybrid energy system is modeled taking into account the hardware availability and unavailability of photovoltaic panels. Various hardware failure events of photovoltaic modules can be represented by a random variable and is assumed to follow a Binomial distribution. In addition, the power output modeling of wind turbine components consider the effect of wind turbine generator force outage rate (FOR). The operating state of wind turbine generator is decided at start of each hour by drawing a uniform random number and finally the sequential up down up cycles of a wind turbine generator are combined with the hourly available wind power to evaluate the actual hourly generated wind power. The design problem of a hybrid energy system can be taken as an optimization problem which maximize or minimize an objective function by changing the values of decision variables. In this study, a composite objective function is formulated fulfilling both economic and reliability criteria subject to various design and operational constraints. The number of each hybrid system component unit such as number of photovoltaic arrays, wind turbine generator, battery storage unit and converter constitute the design variables which values are updated during each objective function evaluation such that the given optimization objective is achieved. The key issue in solving a hybrid energy system design problem is the selection of a sizing technique in terms of computational simplicity and optimal utilization of the available resources. The optimization method adopted in the current study is based on a new meta-heuristic optimization algorithm called as Cuckoo Search (CS) via Levy flights. The developed methodology is implemented in MATLAB programming environment. Unlike other optimization algorithms, the numbers of parameters to be fine tuned during optimization is very less in case of Cuckoo Search (CS) algorithm. The performance of Cuckoo Search (CS) in solving hybrid energy system design problem is tested by comparing with other well known optimization algorithms like Genetic algorithm (GA) and Particle Swarm Optimization (PSO) taking the optimization of three different system schemes such as photovoltaic-battery, wind-battery and photovoltaic-wind-battery applicable to the considered case study area. Analysis result shows better performance of the proposed methodology based on Cuckoo Search (CS) in terms of computational speed and better quality solutions. Moreover, the optimization results indicate that system scheme comprising of photovoltaic, wind and battery storage provides the most techno-economical viable option for meeting the energy requirement of the study area. v | P a g e An analysis on economic performance sensitivity of the optimal hybrid photovoltaic-wind-battery is necessary from the planning perspective point of view. A sensitivity analysis is carried out to determine the impact of various influencing parameters on the cost of energy (COE). The parameters considered are wind speed, solar radiation, capital cost of the system component, load demand, wind turbine generator force outage rate (FOR) and physical unavailability of the photovoltaic panels. Each parameter is varied over a range of values around the base value. The reliability assessment of a renewable energy based hybrid energy system need to be address differently compared to the conventional system of power generation as renewable energy generation units are not considered as fixed capacity generation systems. Several studies have been reported on the analysis of reliability aspects of hybrid energy system which can be broadly grouped into two main categories such as analytical methods and Monte Carlo simulation. Monte Carlo simulation overcomes the limitations of analytical methods by treating the problems as a series of real experiments and simulating the actual stochastic behavior of the system components. Hence, the current work makes use of Monte Carlo simulation for reliability assessment of hybrid energy system through evaluation of loss of load expected (LOLE) reliability index. LOLE is the expected period during which the system load is expected to exceed the available generation capacity. It is expressed as number of hours per year. The reliability analysis is carried out by simulating various contingency conditions such as reduction in solar radiation and wind speed, increase in energy consumption, and increase in failure rate of wind turbine and photovoltaic (PV) panels. The hourly generation from each component of hybrid energy system obtained from the generation models constructed for the purpose is compared with the hourly variation in the load to evaluate the reliability index. Analysis of LOLE evaluation results suggest better reliability performance of photovoltaic-wind-battery system compared to standalone photovoltaic-battery and wind-battery only systems at all the contingency conditions simulated.