The topic of this thesis is the study of energy storage systems operating with wind power plants. The motivation for applying energy storage in this context is that wind power generation is intermittent and generally difficult to predict, and that good wind energy resources are often found in areas with limited grid capacity. Moreover, energy storage in the form of hydrogen makes it possible to provide clean fuel for transportation. The aim of this work has been to evaluate how local energy storage systems should be designed and operated in order to increase the penetration and value of wind power in the power system. Optimization models and sequential and probabilistic simulation models have been developed for this purpose. Chapter 3 presents a sequential simulation model of a general windhydrogen energy system. Electrolytic hydrogen is used either as a fuel for transportation or for power generation in a stationary fuel cell. The model is useful for evaluating how hydrogen storage can increase the penetration of wind power in areas with limited or no transmission capacity to the main grid. The simulation model is combined with a cost model in order to study how component sizing and choice of operation strategy influence the performance and economics of the wind-hydrogen system. If the stored hydrogen is not used as a separate product, but merely as electrical energy storage, it should be evaluated against other and more energy efficient storage options such as pumped hydro and redox flow cells. A probabilistic model of a grid-connected wind power plant with a general energy storage unit is presented in chapter 4. The energy storage unit is applied for smoothing wind power fluctuations by providing a firm power output to the grid over a specific period. The method described in the chapter is based on the statistical properties of the wind speed and a general representation of the wind energy conversion system and the energy storage unit. This method allows us to compare different storage solutions. In chapter 5, energy storage is evaluated as an alternative for increasing the value of wind power in a market-based power system. A method for optimal short-term scheduling of wind power with energy storage has been developed. The basic model employs a dynamic programming algorithm for the scheduling problem. Moreover, different variants of the scheduling problem based on linear programming are presented. During on-line operation, the energy storage is operated to minimize the deviation between the generation schedule and the actual power output of the wind-storage system. It is shown how stochastic dynamic programming can be applied for the on-line operation problem by explicitly taking into account wind forecast uncertainty. The model presented in chapter 6 extends and improves the linear programming model described in chapter 5. An operation strategy based on model predictive control is developed for effective management of uncertainties. The method is applied in a simulation model of a wind-hydrogen system that supplies the local demand for electricity and hydrogen. Utilization of fuel cell heat and electrolytic oxygen as by-products is also considered. Computer simulations show that the developed operation method is beneficial for grid-connected as well as for isolated systems. For isolated systems, the method makes it possible to minimize the usage of backup power and to ensure a secure supply of hydrogen fuel. For grid-connected wind-hydrogen systems, the method could be applied for maximizing the profit from operating in an electricity market. Comprehensive simulation studies of different example systems have been carried out to obtain knowledge about the benefits and limitations of using energy storage in conjunction with wind power. In order to exploit the opportunities for energy storage in electricity markets, it is crucial that the electrical efficiency of the storage is as high as possible. Energy storage combined with wind power prediction tools makes it possible to take advantage of varying electricity prices as well as reduce imbalance costs. Simulation results show that the imbalance costs of wind power and the electricity price variations must be relatively high to justify the installation of a costly energy storage system. Energy storage is beneficial for wind power integration in power systems with high-cost regulating units, as well as in areas with weak grid connection. Hydrogen can become an economically viable energy carrier and storage medium for wind energy if hydrogen is introduced into the transportation sector. It is emphasized that seasonal wind speed variations lead to high storage costs if compressed hydrogen tanks are used for long-term storage. Simulation results indicate that reductions in hydrogen storage costs are more important than obtaining low-cost and high-efficient fuel cells and electrolyzers. Furthermore, it will be important to make use of the flexibility that the hydrogen alternative offers regarding sizing, operation and possibly the utilization of oxygen and heat as by-products. The main scientific contributions from this thesis are the development of - a simulation model for estimating the cost and energy efficiency of wind-hydrogen systems, - a probabilistic model for predicting the performance of a gridconnected wind power plant with energy storage, - optimization models for increasing the value of wind power in electricity markets by the use of hydrogen storage and other energy storage solutions and the system knowledge about wind energy and energy storage that has been obtained by the use of these models. / Paper 1 is reprinted with kind permission of ACTA Press. Paper 2 is reprinted with kind permission of Elsevier/ Science Direct. http://www.elsevier.com, http://www.sciencedirect.com Paper 3 is reprinted with kind permission of IEEE.
Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:ntnu-132 |
Date | January 2004 |
Creators | KorpÄs, Magnus |
Publisher | Norges teknisk-naturvitenskapelige universitet, Fakultet for informasjonsteknologi, matematikk og elektroteknikk, Fakultet for informasjonsteknologi, matematikk og elektroteknikk |
Source Sets | DiVA Archive at Upsalla University |
Language | English |
Detected Language | English |
Type | Doctoral thesis, comprehensive summary, info:eu-repo/semantics/doctoralThesis, text |
Format | application/pdf |
Rights | info:eu-repo/semantics/openAccess |
Relation | Doktoravhandlinger ved NTNU, 1503-8181 ; 2004:39 |
Page generated in 0.003 seconds